Human‐induced salinity changes impact marine organisms and ecosystems

Climate change is fundamentally altering marine and coastal ecosystems on a global scale. While the effects of ocean warming and acidification on ecology and ecosystem functions and services are being comprehensively researched, less attention is directed toward understanding the impacts of human‐driven ocean salinity changes. The global water cycle operates through water fluxes expressed as precipitation, evaporation, and freshwater runoff from land. Changes to these in turn modulate ocean salinity and shape the marine and coastal environment by affecting ocean currents, stratification, oxygen saturation, and sea level rise. Besides the direct impact on ocean physical processes, salinity changes impact ocean biological functions with the ecophysiological consequences are being poorly understood. This is surprising as salinity changes may impact diversity, ecosystem and habitat structure loss, and community shifts including trophic cascades. Climate model future projections (of end of the century salinity changes) indicate magnitudes that lead to modification of open ocean plankton community structure and habitat suitability of coral reef communities. Such salinity changes are also capable of affecting the diversity and metabolic capacity of coastal microorganisms and impairing the photosynthetic capacity of (coastal and open ocean) phytoplankton, macroalgae, and seagrass, with downstream ramifications on global biogeochemical cycling. The scarcity of comprehensive salinity data in dynamic coastal regions warrants additional attention. Such datasets are crucial to quantify salinity‐based ecosystem function relationships and project such changes that ultimately link into carbon sequestration and freshwater as well as food availability to human populations around the globe. It is critical to integrate vigorous high‐quality salinity data with interacting key environmental parameters (e.g., temperature, nutrients, oxygen) for a comprehensive understanding of anthropogenically induced marine changes and its impact on human health and the global economy.


| INTRODUC TI ON
Climate change is transforming the world as we know it, driving fundamental changes to the physical climate and subsequently impacting biodiversity, ecosystem functioning, and our socioeconomic livelihood (IPCC, 2023). In the marine and coastal biosphere, the current body of research on the ecological impact of climate change primarily addresses the effects of ocean warming and ocean acidification (Doney et al., 2012;Henson et al., 2017). By comparison, salinity shifts due to the changing water cycle (Durack et al., 2012) and its hydrological extremes (Westra et al., 2014), melting ice (Haumann et al., 2016), and altered terrestrial runoff (Lenderink & Van Meijgaard, 2008) have received less attention, similar to ocean deoxygenation , despite its considerable impact on marine biota, ecosystems, and human livelihood (Newell, 1976).
Here, we highlight the central role that salinity plays in the context of human-induced environmental change both for current and future climate change scenarios and show that for certain ecosystems, salinity changes are the leading physical driver of biological response.
We explore interactions between key environmental parameters that often accompany salinity changes, such as temperature, oxygen solubility, nutrient concentration, and sedimentation. We discuss both long-and short-term salinity changes and how they will shape the evolution of oceans, its ecosystems, and inhabitants.

| SALINIT Y-AK E YVARIAB LEINA CHANG INGENVIRONMENT
Open ocean and coastal salinity are driven by water cycle variability and change that modulate the physical ocean system. The ocean is responding to the dual forcings of surface-driven warming and an enhanced atmospheric water cycling at its interface and boundaries (cryosphere and terrestrial runoff) (Durack, 2015). The complex interactions in response to these dual forcings are modulating surface ocean properties, which are subducted into the ocean interior (Durack & Wijffels, 2010). For example, coastal and estuarine ecosystems are strongly impacted by freshwater availability from rainfall and the associated runoff, with sensitivities to extreme flood and drought events (Gibson et al., 2002). For the global ocean, salinity is a marker for water cycle change, with the ocean integrating changing freshwater fluxes from the atmosphere, cryosphere, and land surface. A clear and robust pattern of salinity change, demarking changes to the oceanic water cycle, has been observed since the 1950s and since then reproduced by key studies (Stocker et al., 2013). To date, we have already observed changes in ocean salinity at an order of 0.05 to 0.01 per decade on the Practical Salinity Scale (PSS-78) at an accelerating rate (~0.2 to ~0.35 per 50 years (Durack et al., 2012); Figure 1a). These changes indicate that the open ocean is following the global idealized 'wet gets wetter, dry gets drier' pattern of water cycle change caused by increasing atmospheric temperatures (Chou & Neelin, 2004;Held & Soden, 2006). Fresh regions accordingly appear to becoming fresher and salty regions saltier, evident in the ocean near-surface upper layers and extending through intermediate depths (Durack et al., 2012;Durack & Wijffels, 2010). For the near surface, salinity increases in the subtropics (10-40°N/S) and freshens in the tropics (0-10°N/S) and subpolar and polar regions (40-90°N/S). Changes at depth generally follow the same patterns, with salinity changes following the subduction and circulation pathways into the ocean interior (Durack & Wijffels, 2010).
Future projections using the previous CMIP5 model suite (Taylor et al., 2012) are indicating these currently observed changes are likely to intensify (Stocker et al., 2013). By the end of the century, near-surface changes are projected to increase markedly, by an order of ~0.5 to 1.0 PSS-78 (Stocker et al., 2013), and in the deep by about 0.05 PSS-78 (Heuzé et al., 2015) (Figure 1b-d). Their magnitude is proportionally linked to the rate of global mean surface air temperature increase, with larger atmospheric CO 2 concentrations driving stronger warming and salinity responses (Stocker et al., 2013;Tebaldi et al., 2021). Water mixing through ocean currents and in particular the thermohaline circulation (THC) is largely salinity (and temperature) driven, with regional changes resulting from even small perturbations driving changes to Earth's climate directly (Carmack et al., 2016). For example, ocean freshening from melting glaciers, ice sheets, and other terrestrial water, particularly in coastal regions, is contributing to enhanced ocean stratification (Li et al., 2020).
Continued high latitude terrestrial freshwater input seems to contribute to slowing of the THC (Boers, 2021). in dynamic coastal regions warrants additional attention. Such datasets are crucial to quantify salinity-based ecosystem function relationships and project such changes that ultimately link into carbon sequestration and freshwater as well as food availability to human populations around the globe. It is critical to integrate vigorous highquality salinity data with interacting key environmental parameters (e.g., temperature, nutrients, oxygen) for a comprehensive understanding of anthropogenically induced marine changes and its impact on human health and the global economy.

K E Y W O R D S
algae, coastal, coral reef, deep-sea, ecosystem services, mangrove, plankton, polar, seagrass, tidal marsh While broad-scale, open ocean salinity changes are coherent and physically consistent at basin scales across observations and models, the coastal regions where ocean and land processes intersect, are more dynamic in their response to climate change. Coastal regions and estuaries are among the most productive ecosystems on Earth (Nixon et al., 1986). However, due to their close proximity to human settlements, they are arguably more directly affected by local anthropogenic impacts (e.g., land use, urbanization, drainage, river regulation and discharge, changes in vegetation, desalination discharge, and terrestrial runoff) compared to the open ocean.
Concomitantly, climate change-driven water cycle amplification leads to enhanced rainfall, which enhances flooding and runoff in some parts of the globe, while in other regions its decrease is causing droughts and increased evaporation (Masson-Delmotte et al., 2021). As a result, terrestrial discharge is changing dramatically in some parts of the world (Masson-Delmotte et al., 2021), a trend projected to continue in the future (Figure 1e,f). Particularly, local anthropogenic impacts and increasing precipitation extremes, along with enhanced drying and evaporation can result in concerted pulses of terrestrial runoff, which can strongly affect coastal ecosystems (Lenderink & Van Meijgaard, 2008;Papalexiou & Montanari, 2019). However, temporal and spatial data characterizing such changes as a baseline for large-scale coastal impact assessment are currently lacking.
Coastal ecosystems are also exposed to additional stressors associated with or caused by salinity changes, for instance salinity levels together with temperature drive water density. In cold-water polar oceans, a small change in salinity results in density changes that would require a far larger temperature change, while conversely, temperature is the leading property for most of the tropics and subtropics (Griffies et al., 2014). Water density drives oxygen saturation levels (Figure 2), and freshwater stratification and nutrient driven microbial respiration can additionally trigger coastal hypoxic areas (Breitburg et al., 2018).

BOX1:Keytermsanddefinitions
An ecotone is a transition area where two (or more) communities meet and integrate.
Harmful algal blooms (HAB) occur when algal populations reproduce exponentially and produce toxic or harmful effects on humans and/or marine organisms.
A holobiont encompasses a host organism and all of its associated microorganisms that directly or indirectly affect the resultant metaorganism's health and functioning.
Hydrological cycling represents the flow of freshwater through Earth's Systems, including the ocean, cryosphere, atmosphere, and land in all phases (i.e., solid (ice), liquid, and vapor).
Microbiome is the sum of microorganisms, including viral, prokaryotic, and microeukaryotic members and their genetic material.
Microbiomes can be free-living or associated with non-living substrates (soil, sediment) and living hosts.
Based on osmotic stress, osmoacclimation refers to (reversible) biochemical, biophysical, and physiological modifications in cellular structure and/or function, while osmoadaptation refers to genotypic changes.
Practical-Salinity-Scale 1978 (PSS-78) is a defined, dimensionless salinity scale based on electrical conductivity, which is influenced by the ambient water temperature and pressure.
Salinization is the increase in salt concentration or salt accumulation. Salinization can occur in soils or water/porewater. Sea level rise (SLR) is an increase in the level of the world's oceans due to the effects of climate change. This includes thermal water expansion related to increasing ocean temperatures (thermosteric effect), mass/water addition from melting ice reservoirs and transfer of terrestrial (surface and subsurface) water to the ocean (eustatic effect), and changes in salinity (and thus water density, with an expansion for freshening and contraction for salinization; halosteric effect).

Sea surface salinity (SSS)
is the amount of dissolved salts in the near-surface layer of the ocean-atmosphere interface. Concentrations are measured in situ as well as by satellites since 2010. Changes are driven by evaporation and precipitation patterns and ocean circulation in the open ocean, plus terrestrial runoff in the near-coastal zones.
A thermocline/halocline is a distinct aquatic layer in which temperature/salinity changes occur much stronger with depth than in the layers above or below. The strong gradient separates the well-mixed ocean (above), which is in equilibrium with the atmosphere directly above the ocean surface, and the ocean interior (below). Changes to both state variables may strongly contribute to ocean stratification.
The thermohaline circulation (THC), a global current, is driven by temperature (thermo) and salinity (haline) related differences in water density.
fundamentally transform coastal ecosystems. These stressors are all predicted to increase as a result of climate change and anthropogenic activity (Knight & Harrison, 2009;Sinha et al., 2017). Ecosystems at the coastal-terrestrial interface are also increasingly affected by local sea level rise (SLR) (Nicholls, 2010). SLR is driven by thermal expansion (thermosteric effect), freshwater addition (eustatic effect), and salinity distribution changes (halosteric effect) Munk, 2003) (Figure 2). As a result, coastal and low-lying inland wetlands, islands, and estuaries are experiencing saltwater intrusion at the terrestrial-ocean interface. This salt ion increase and salinization affect ecosystem structure and function, including comprehensive changes in biogeochemical cycles for example (Herbert et al., 2015).

| EFFEC TSOFSALINIT YONTHE G LOBALO CE ANECOSYS TEM
Salinity is a key variable shaping the distribution of biota across all trophic levels in marine ecosystems (Kirst, 1990;Newell, 1976) ( Figure 2). In the following, we review the impacts of salinity for three key marine biomes-open ocean, polar, and coastal-due to their vulnerability to salinity changes and the breadth of impact to ecological keystone species, ocean productivity, and humans.
We also outline details of the ecophysiological effects of selected keystone species in Table 1. Given the profound impact of salinity on microbiomes and their impact on downstream ecosystem structuring (Lozupone & Knight, 2007), we first consider changes across the world's ocean at large, before diving into the three key biomes.
We then assess open ocean salinity effects on epipelagic and deep pelagic systems that profoundly differ in their physicochemical and biological constitution. Largely well-connected with the open ocean, we next explore polar regions that are driven by annual freezing and brine rejection cycles that fundamentally govern the ecophysiology of its inhabitants, in particular with regard to salinity (Ardyna & Arrigo, 2020). We then focus on coastal ecosystems and highlight how changes in salinity patterns affect benthic subtidal coral reefs, benthic intertidal estuaries, and the coastal photoautotrophs in the littoral zone. With their pivotal roles in ecosystem services related to carbon sequestration and storage, nutrient cycling, biodiversity, fisheries, and coastal protection, F I G U R E 1 Changes in global salinity patterns and terrestrial runoff. Due to an intensified water cycle, global surface salinity patterns (a) have changed from 1950 to 2020 (PSS-78 70 years). Future projections from the phase 6 of the Coupled Model Intercomparison Project (CMIP6) multi-model mean under (b) low (Shared Socioeconomic Pathway, SSP1-2.6), (c) mid (SSP3-7.0), and (d) high (SSP5-8.5) CO 2 emission future scenarios (differences from the 1985-2014 historical counterpart) show strong salinity changes (PSS-78) that resemble the pattern of observed changes will continue, particularly under mid and high CO 2 emission scenarios. This is in concert with changes in precipitation and more direct anthropogenic impact, changes over land annual total runoff (%), that project increasing variations from (e) low (SSP1-2.6), (f) mid (SSP3-7.0), and (g) high (SSP5-8.5) CO 2 future emission scenarios. The CMIP6 Earth Systems Models (ESMs) that provide salinity and runoff change fields are detailed in the Supplementary Material & Methods (Tables S2-S4). Salinity is a major physicochemical factor determining bacterial community composition, and salinity differences pose a particularly strong barrier for microorganismic metabolism to overcome (Jurdzinski et al., 2023). Salinity drives soil and water microbiome global composition for both, prokaryotic and eukaryotic taxa (Herlemann et al., 2011;Lozupone & Knight, 2007;Shearer et al., 2007;Telesh et al., 2013). Given the significant effect of salinity on the microbiome at such large scales (Lozupone & Knight, 2007), future shifts in salinity due to ice cap melting, precipitation variability, and SLR has the potential to alter global biogeochemical cycling, the effects of which are likely compounded by further climate change (Cavicchioli et al., 2019). Teasing apart the functional and metabolic shifts of the microbiome (or key members within) will be key to our ability in predicting biogeochemical and biological consequences of salinity change. Accordingly, biodiversity shifts are projected as a result of salinity changes due to diversification within new niches (Herbert et al., 2015;Shearer et al., 2007;Telesh et al., 2013) ( Figure 3). Although many cosmopolitan microbial taxa will persist throughout regional salinity changes, their function may change (Herbert et al., 2015;Telesh et al., 2013). Additionally, salinity changes will affect key carbon-and nitrogen-cycling functional groups across all ecosystems (Table 1). For both carbon and nitrogen cycles, the quality and availability of organic carbon through sedimentation, litter production and quality, and exudates will affect what the microorganisms can degrade (enzymatic repertoire (Balmonte et al., 2020;Celussi et al., 2019)), and the efficiency at which they can do this (biogeochemical pathways (Luo et al., 2019;Neubauer et al., 2013;Santoro, 2010)). The effect of increasing salinity changes on the carbon cycle will largely depend on the organic matter-microbiome relationship, with predictions of both enhanced and suppressed carbon sequestration (Luo et al., 2018;Neubauer et al., 2013). It is further predicted that seawater intrusion and regional salinity increase will F I G U R E 2 Conceptual diagram highlighting salinity's central role in structuring the physical and biological environment under human-induced environmental change. Effects of salinity are depicted in dark blue. Adapted from Doney et al. (2012).

TA B L E 1
Overview on selected key ecophysiological effects of salinity changes for organisms within selected marine and coastal ecosystems (for references and additional considerations see Table S1).

Organism (ecosystem) Physiological effect Ecological consequence(s) Potential socioeconomic impact
Microorganisms [soil, sediment, and water column; holobionts (host-associated microorganisms)] Changes in enzymatic capacity and metabolic rates. Shifts in community composition. Changes in holobiont functioning either directly through association with different microbes, or indirectly, through physiological changes of the holobiont (e.g., the osmolyte metabolism). The microbiome contributes to salinity tolerance of the host.
Shift(s) in the type and rate of organic matter degradation, which will affect soil stability and accumulation. Reduced diversity, favoring euryhaline holobionts/species. Shift(s) in the type and rate of organic matter degradation and nutrient availability will affect greenhouse gas emissions for both, the rate and the type of greenhouse gas production (CH 4 , CO 2 , and N 2 O).
Coral (coral reef) Changes in respiration/photosynthetic rates, reduced/heightened coral bleaching/mortality susceptibility for substantial salinity changes, lowered resilience due to low level chronic osmotic stress, changes in associated microbial community composition (pro-and eukaryotic), interactive effects with temperature changes, increased larval and in particular gamete mortality for considerable salinity drops.
Loss of coral diversity, loss of ecosystem, loss of associated diversity (coral reefs harbor estimated 30% of marine species).
Loss of ecosystem functions and value (e.g., food source, coastal protection, and tourism).

Mangroves and marshes (coastal)
Reduction in growth and seedling survival and development, and increased mortality due to elevated salinity. Ionic or sulfide stress leading to reallocation of resources, for example, reduction in photosynthesis, may be exacerbated by other stressors, such as elevated temperature and drought. Short-term hyposalinity pulses may benefit coastal plants, including seedling success and range expansion seaward if sedimentation allows; however, long-term freshwater conditions can lead to reduction in growth or die-backs.
Reduced diversity, favoring fast-growing and adaptable taxa. Ecotone shifts, favoring halotolerant plants, or rstrategists (e.g., algae-dominant or herbaceous over woody taxa), with broad consequences to ecosystem function, trophic cascades, and lignocellulose deposits from detritus. Reduction of autochthonous carbon inputs from litter and exudates.
Estuary relocation; loss of ecosystem functions and value (e.g., ecosystem services to coral, seagrass, plankton, fish; loss of freshwater source).

Phytoplankton (coastal, open ocean, and polar regions)
Changes in respiration/photosynthetic rates and disruption of cellular processes. Toxin(s) release by harmful algal species.
Water column stratification promotes cyanobacterial success and challenges diatoms. Halotolerant and euryhaline HABs may be favored and impact ecosystem structure and function. Shifts in phytoplankton communities may trigger trophic cascades altering species composition on higher trophic levels, particularly in polar regions.
Salinity changes can trigger/ facilitate phyto-and/or bacterioplankton HABs that particularly impact shellfish, fish, marine mammal, and bird population, which in turn directly impact human health, food supply, and economy. Reduced fish stocks affected by trophic cascades.
Macroalgae (coastal) Changes in respiration/photosynthetic rates and disruption of cellular processes.
Negative effects on sexual reproduction (e.g., lowered gamete viability, fertilization, and polyspermy) and altered disease susceptibility can lead to decreased genetic diversity that may change ecosystem structure and function.
Loss of ecosystem functions and value (e.g., ecosystem services of coral, seagrass, plankton, fish, and tidal marshes). cause a shift from methanogenesis-dominated to sulfate reductiondominated metabolism, which will likely lower CH 4 production in the short term (Neubauer et al., 2013). For the nitrogen cycle, projected increases in nitrogen mineralization and reduced coupled nitrification-denitrification could result in increased NH 4 + export from groundwater systems (reviewed in Santoro, 2010).
Besides affecting free-living microorganisms, salinity changes can alter host-associated microbiome structure and therefore holobiont functioning. This can occur either directly through association with different taxa (e.g., Röthig et al., 2016;Yuan et al., 2016) or indirectly through physiological changes of the holobiont, such as osmolyte metabolism (Gegner et al., 2017;Ngugi et al., 2020;Ochsenkühn et al., 2017). Marine organisms in general tend to have a broad salinity tolerance, potentially due to their exposure to varying salinity levels over seasons or as a result of oceanic cycles (Velasco et al., 2019). At the same time, such altered salinity levels impact organism physiology and can affect their stress tolerance, in particular with regard to temperature stress (Gegner et al., 2017). At large, salinity has been shown to drive community structure, highlighting the ecological importance of changing salinity levels (Coles & Jokiel, 1992;Lozupone & Knight, 2007;Wilson & Dunton, 2018).

| Openocean
In the open ocean, salinity in combination with light, temperature, nutrients, circulation, carbon dioxide, and oxygen determine the physiology of life in this biome, ultimately impacting ecosystem composition, structure, and functioning. Salinity therefore defines plankton ecological niches, affecting growth, photosynthesis, and biogeochemical activities and thus diversity (Bijma et al., 1990;Dueñas-Bohórquez et al., 2009;Mezger et al., 2016;Qasim et al., 1972) Guidi et al., 2016). More recently, salinity has even been linked to plankton predicted ecological interactions topologies (Chaffron et al., 2021). For example, a Trades-like plankton community ( Figure S1) has been predicted to be vulnerable to salinity changes at tropical latitudes (Chaffron et al., 2021), where changing salinities (+~0.5-1.0 PSS-78 by the end of the century (Figure 1a,b)) are well able to profoundly impact community structure (Figure 4a). Loss of ecosystem functions and value (e.g., ecosystem services of coral, seagrass, plankton, fish, and tidal marshes).
Fish, invertebrates (e.g., mussels, echinoderms, crustaceans) (coastal and polar regions) Disruption of cellular processes and metabolic rates (e.g., caused by enhanced energetic requirements for osmo-adjustments) are common in marine fish and invertebrates and may disturb fertilization, development, and sensory perception, increase susceptibility to toxins/pollutants, and lead to changes in respiration, feeding rates and size/growth.
Reduced diversity favoring euryhaline species can result in trophic collapse (e.g., echinoderms, jellyfish, copepods, and fish) due to bottom up and/ or direct impact leading to loss of ecosystem structure and function.
Loss of ecosystem functions and value, for example water quality degradation or loss of food security from fishing and/or aquaculture.

F I G U R E 3
Overview of salinity changes and their influence on ecosystem function. Increasing temperatures enhance hydrological cycling, resulting in increased meltwater and shifts in evaporation and precipitation patterns that affect global ocean salinity patterns. Local anthropogenic impacts, including land use practice further modulate terrestrial runoff patterns affecting coastal ecosystems. Enhanced variation or shifts in salinity impact diversity, growth, and survival of key species. Sea level rise connected with salinization as well as ecotone shifts and trophic cascades may contribute to substantially altered ecosystem structure and functionality. data in this area stem from inferences that are acquired from betterstudied systems such as the polar regions.

| Polarregions
In supply, and primary production (Carmack et al., 2016). For example, polar ecosystems have experienced comprehensive community structure changes (Greene & Pershing, 2007) in phytoplankton biomass and production (Blais et al., 2017). Rising temperatures and the reduction in sea ice extent have prolonged the phytoplankton growing season, thus increasing net primary production, and seem to favor photosynthetic picoeukaryotes over diatoms (Ardyna & Arrigo, 2020;Blais et al., 2017). While terrestrial runoff supplies nutrients that facilitate plankton growth, salinity-enhanced stratification may at least regionally suppress plankton biomass and productivity (Ardyna & Arrigo, 2020). As primary consumer (e.g.,

| Coastalecosystems
Salinity changes can have a profound influence on the distribution and physiological responses of diverse multicellular organisms

| The benthic subtidal-Coral reefs
Research on the effects of salinity changes and its implications for corals and reef ecosystems is relatively scarce, despite its established importance with respect to osmoadaptation (sensu Reed, 1984) and holobiont physiology (Coles & Jokiel, 1992). This may be due to the circumstance that 'historically' coral reefs occur in regions where salinity has been relatively stable. In addition, ongoing research largely focuses on the more imminent threat associated with ocean warming (i.e., coral bleaching) (Dietzel et al., 2020). However, salinity is considered a key environmental variable for coral distribution, and salinity changes can also lead to coral bleaching and mortality (Coles & Jokiel, 1992). For this reason, it is important to consider how salinity, either in isolation, or in combination with ocean warming and the dynamical effect of the ocean physics, along with ocean acidification will affect corals and the reef ecosystems they build. We therefore applied distribution modeling to explore  (Coles & Jokiel, 1992). Physiological responses (e.g., osmoacclimation sensu Reed, 1984) of corals to both increasing and decreasing salinity can vastly differ and some species seem more osmotolerant than others (Coles & Jokiel, 1992). In general, corals seem to cope well under high salinity conditions (e.g., ~40 PSS-78 in the Red Sea and the Persian/Arabian Gulf) and even appear to exhibit decreased bleaching susceptibility (D'Angelo et al., 2015;Ochsenkühn et al., 2017). The presumed increased thermal tolerance is supposedly mediated by osmolytes that harbor antioxidative functions and can thus counter reactive oxygen species (ROS) from heat stress (Gegner et al., 2019;Hume et al., 2016;Osman et al., 2017). By comparison, decreased salinity seems to have rather detrimental effects, as evidenced by the many studies describing reduced respiration, photosynthetic rates, and mortality (critical threshold ~20 PSS-78 for 24 h) consequential to lowered salinity levels due to heavy rainfall associated with major storms or monsoons (Coles & Jokiel, 1992;Ferrier-Pagès et al., 1999;Moberg et al., 1997) (Table 1). This may explain why the effects of decreased salinity are better represented in the scientific literature than the effects of salinity increase. While quantitative data are lacking, such short-term salinity variations may well pose a larger pressure on coral than the more consistent, longterm ocean scale salinity changes (Figure 1a,b). Taken together, even though salinity is considered a key variable for coral distribution and health, interaction effects with other stressors (e.g., ocean warming, ocean acidification, ocean deoxygenation), cellular response mechanisms (Mayfield & Gates, 2007), or the role of associated microorganisms (Röthig et al., 2016), which at least partially affect future distribution ranges, are currently unclear. This is surprising given the ecological and economic importance of tropical coral reefs matching or even surpassing estuaries. However, the latter seem to affect industrialized nations more directly, which may partially explain why salinity effects in estuaries are better understood.

| The benthic intertidal-Estuary continuum
The effect of salinity on mangrove and coastal/freshwater tidal marsh ecosystems in terms of physiological, reproductive, and ecological stress and adaptations has been well-summarized in several reviews over the past 10 years (e.g., Herbert et al., 2015;Parida & Jha, 2010;Pratolongo et al., 2018;Wang et al., 2011). Therefore, here we will focus on the overarching effects of salinity changes along the mangrove-marsh continuum and in the context of precipitation and salinization (Figures 1 and 3, Table 1). In contrast to coral reefs, low salinities induced by regular or relative increases in precipitation will largely benefit coastal wetlands (Figure 3). This is primarily due to the increased access to groundwater, thereby enhancing growth and biodiversity (Feher et al., 2017;Gilman et al., 2008;Janousek & Mayo, 2013); however, chronic freshwater conditions will lead to the eventual reduction or die-off of tidal marsh and mangrove plants (Wang et al., 2011) (Table 1). In instances where runoff is concomitant with increased precipitation, mangrove expansion both seaward, landward, and up-estuary will be facilitated in the long term by increased sedimentation, sediment volume, and inundation (Gilman et al., 2008;Yang et al., 2013). Additionally, some coastal tidal marsh species can grow under meso-and oligo-haline conditions, suggesting certain taxa, particularly those adapted to high marsh conditions, could migrate to fresher ecotones under SLR (Janousek & Mayo, 2013;Kirwan & Gedan, 2019;Sutter et al., 2015) ( Figure 3).
Conversely, salinization caused by SLR, drought, and low tidal flushing will ultimately have a negative impact on this ecosystem and/or disrupt the health and function of emergent coastal autotrophs (Table 1). Along the estuarine continuum, freshwater tidal and upland ecosystems may be the biggest 'losers' connected to changes in salinity. Encroachment of halophyte species into fresher areas as saltwater reaches upstream may initially cause diversification of species at the ecotone and be an advantage to brackish taxa. However, these changes will ultimately be at the expense of freshwater tidal ecosystems (Kirwan & Gedan, 2019). Salinity shifts not only lead to poor competition of freshwater tidal species, but salinity and inundation stress may also create a positive feedback loop for tidal wetlands at large, whereby stress and low reproduction leads to increased plant death and decay and decreased net carbon inputs, which then cause soil compaction and subsidence (Herbert et al., 2015). The indirect salinity-induced subsidence and predicted increase in inundation would result in more stress to the ecosystem due to a loss of viable habitats, leading to less autochthonous carbon inputs into the system (Herbert et al., 2015). Salinization can also negatively impact mangrove and coastal marshes in the short-tomedium term via ion and sulfide stress (e.g., drought, low tidal flushing, high evapotranspiration (Gilman et al., 2008;Luo et al., 2019)), which can eventuate to large-scale die-backs (e.g., Alber et al., 2008;Lovelock et al., 2017). However, salinization in the context of relative SLR does show promise for coastal wetlands in scenarios where sedimentation and elevation provide new habitat and carbon sink opportunities for mangroves and tidal marsh ecosystems . These contrasting impacts highlight the complexities of predicting interactive effects of salinity changes across multiple spatiotemporal scales and under future climatic conditions.

| The littoral zone
Ocean productivity largely refers to the production of organic matter by phytoplankton or photoautotrophs. Often distinct from the open ocean, productivity in coastal ecosystems is driven by proximity to land and its nutrient sources, the interception of organic matter by the shallow seafloor, and the propensity for coastal upwelling, all resulting in highly productive ecosystems. The response of phytoplankton to salinity changes is therefore of key importance and comparably well researched (Bijma et al., 1990;Dueñas-Bohórquez et al., 2009;Mezger et al., 2016;Qasim et al., 1972) (Table 1).
However, a key challenge in predicting how coastal phytoplankton will respond to changes in salinity is drawing meaningful conclusions from the high degree of inter-and intraspecific response variability that has been reported in both laboratory and field studies (including species associated with harmful algal blooms (HAB)). These responses may be influenced by any combination of physiological osmoacclimation and phenotypic plasticity (i.e., variation within strains), genetic variation among strains, and evolutionary osmoadaptation (Burford et al., 2020). Despite this challenge, it is anticipated that the most pronounced changes in phytoplankton diversity and community structure will occur along ecotones or ecoclines of the freshwater-marine continuum (Mazzei & Gaiser, 2018;Muylaert et al., 2009;Quinlan & Phlips, 2007) (Figure 3). Particularly, genera of freshwater toxigenic phytoplankton are capable of adapting to new conditions and have already been reported in estuaries throughout North America, Australia, Europe, and Japan (Preece et al., 2017). In certain cases, HABs may be induced by salinity changes and some species can even develop enhanced toxicity under altered salinities (Fu et al., 2012). Unfortunately, there continues to be a considerable gap in our understanding of the adaptive mechanisms employed by toxic phytoplankton in response to salinity-induced stress. While some species/strains can adapt and modulate toxin production, others succumb to osmotic thresholds resulting in cell lysis and toxin release into the environment (des Aulnois et al., 2019;Ross et al., 2019;Tanabe et al., 2018).
Similar to phytoplankton, macroalgal species composition and richness can change dramatically along salinity gradients that are typically encountered in transitional coastal regions. Of particular concern are macroalgal populations that exist at the salinity range margins of a species' distribution. For example, climate changerelated declining salinities in the Baltic Sea (currently ranging from 2 to 13 PSS-78 (Kniebusch et al., 2019)) are predicted to continue and push the distribution of marine species on a southerly trajectory, transitioning the community toward a more limnic state (Takolander et al., 2017). Such increases in hyposalinity have the potential to negatively affect macroalgal sexual reproduction by reducing gamete viability and fertilization and promoting polyspermy (Serrão et al., 1996;Steen, 2004). The ensuing shift to asexual reproduction promotes lower genetic diversity and may consequently impair the potential for adaptation to new selective marine regimes (Rothäusler et al., 2018) (Figure 3). Furthermore, increasingly hyposaline conditions of the Baltic Sea may also cause a shift toward the prevalence of green algae (Chlorophytes) with a concomitant decline in red (Rhodophyta) and brown algal (Phaeophyceae) species (Nielsen et al., 1995;Ojaveer et al., 2010;Snoeijs, 1999). This has the potential to affect higher trophic levels, as many green algae in this region are considered to be fast-growing, opportunistic annual species unable to provide appropriate habitat complexity for fauna compared to larger perennial taxa (Takolander et al., 2017).
Seagrasses are another photoautotroph that will be impacted by regional shifts in salinity. In subtropical seagrass beds, changes in salinity can cause a phase shift to occur from dominance of benthic rhizophytic algae to drift algae at sites experiencing lower salinity (Biber & Irlandi, 2006) (Figure 3). While rhizophytic algae can help facilitate sediment stabilization and promote seagrass succession (Thayer et al., 1994;Williams, 1990), large clumps of detached drift algae often reduce seagrass productivity and biomass via the process of over-shading (Cambridge et al., 1986;Hauxwell et al., 2001).
Interestingly, despite the broad geographic range covered by seagrasses (found on all continental margins except the Antarctic), a large number of these coastal and estuarine plants have optimal growth rates between salinities of 30-40 PSS-78 (Touchette, 2007).
Most species can tolerate short-term salinity fluctuations. However, prolonged exposure to hypo-or hypersaline conditions can have a significant impact on plant fitness with subsequent ecosystem ramifications (Collier et al., 2014;Robblee et al., 1991;Wilson & Dunton, 2018) (Table 1, Table S1, Figure 3). Altered salinity can not only have a direct effect on seagrass communities by affecting plant physiology (i.e., disruption of cellular processes), but can also have an indirect effect by modifying the rate of top-down interactions with grazers (Bell et al., 2019). Climate change models project decreased precipitation and increases in the occurrence of drought conditions in such regions as the Mediterranean Sea, the Gulf of Mexico, and southwest Australia (Stocker et al., 2013) (Figure 1e-g). Given that these areas support extensive seagrass meadows, plant sustainability will depend on the ability to adapt to the predicted increased saline conditions (Green et al., 2003). Hypersalinity has already been implicated as a major factor in large-scale die-offs of seagrass in the United States and southern Australia (Robblee et al., 1991;Seddon et al., 2000). As wetland hydrology and seasonal rainfall patterns continue to change, die-offs in these areas may very well increase in frequency and coverage (Duffy et al., 2019). This is of particular concern for vulnerable and near-threatened seagrass species and their associated fauna (Short et al., 2011). Other seagrass populations will have to contend with contrasting hyposaline conditions, occurring in response to enhanced rainfall, runoff, or other hydrological modifications (Figure 1b-g). For example, salinity in the Baltic Sea is expected to decrease by 8%-50% by the end of the 21st century (Meier, 2006).  (Table 1, Table S1, Figures 3 and 4). For some ecosystems, salinity is even considered a key variable to quantify degradation (Bland et al., 2017), but knowledge on the direct effects of salinity changes on marine ecosystem function remains fragmentary.
For example, the loss of ecosystem services (such as nutrition, coastal protection, and tourism) associated with coral reefs would equate to an annual loss of US$375 billion (Costanza et al., 1997).
Losing seagrass meadows, marsh grass and mangroves ecosystems, and their economically valuable species would result in an annual loss of US$22-31 million in carbon value in Australia alone (Serrano et al., 2019). Although it is likely that isolated shifts in salinity would not be the major driver of demise in these ecosystems, healthy ecosystem function is impaired by salinity changes, especially when concurrent with other stressors (e.g., ocean warming, ocean acidification). Indeed, increasing SLR has been estimated to directly threaten an estimated 0.2%-4.6% of the human population that reside in coastal regions with expected annual losses of 0.3%-9.3% of the global gross domestic product by 2100 (Hinkel et al., 2014). For instance, salinity intrusion in soils and freshwater reservoirs, partly connected to annual flooding, already impacts agricultural productivity globally and is a major driver for human migrations (Chen & Mueller, 2018). SLR and changes in terrestrial runoff patterns further affect salinity regimes in estuaries, including upstream salinization and water quality degeneration, thus limiting freshwater supply for the often highly populated coastal regions (Figure 3). Furthermore, other salinity-driven shifts in ocean productivity, such as HABs and/ or anoxic dead zones, can decimate local and regional fish stocks with direct and severe consequences for fisheries, aquaculture, and human health (estimated US$8 billion global annual loss from HABs alone ). Aquaculture may further be directly impacted, exemplified by the economically important mussel genus Mytilus with a global distribution (>750,000 tons production in 2019 (FAO, 2021)).
Here, decreasing salinity can cause reduced growth rates (Riisgård et al., 2013), while increasing salinity variability can affect the physiology and immunology of mussels with further anticipated economic downstream effects (Bussell et al., 2008;Duarte et al., 2018 (Gunderson et al., 2016). However, for many important groups these data are also lacking, exemplified by bacterial (Baker-Austin et al., 2013) and viral (Brown et al., 2009) pathogens, the spread and infectivity of which can be affected by salinity.
Existing large-scale sampling and monitoring efforts (e.g., plankton (Biller et al., 2018) and seagrass (Short et al., 2011)) should consider salinity as a central physio-ecological driver. Future research should also characterize large-scale spatiotemporal coastal salinity patterns and changes therein as a framework to integrate meaningful data such as key species' tipping points (see our modeling approaches for plankton communities and coral distribution based on open ocean salinity projections for two examples of how this can be undertaken (Figure 4)). If conducted, such data can deliver meaningful projections to inform conservation and management strategies.

| CON CLUS ION
The physiological effects of salinity changes in the marine ecosphere have long been studied. However, in the context of climate change, salinity appears to be the 'elephant in the room'. This is particularly remarkable as there is ample evidence that ocean warming and ocean acidification directly affect and are affected by salinity changes. More work is therefore needed to integrate how all three climate change-related stressors interact and affect marine ecosystems and biota, which should be comprehensively summarized as a first step. While physiological (and mostly laboratory-based) data on salinity variations are available for a range of organisms, we are missing baseline data on the natural variability of many ecosystems, which makes any prediction of the consequences of climate change driven salinity changes challenging. In particular, salinity patterns in coastal ecosystems are lacking spatial and temporal resolution.
We also need to acquire a clearer understanding of interactions of salinity with accompanying factors such as temperature, acidification, oxygen solubility, sedimentation, and nutrient input-all of which play a crucial global role for human health and economy.

CO N FLI C TO FI NTE R E S TS TATE M E NT
The authors declare no conflict of interest.

DATAAVA I L A B I L I T YS TAT E M E N T
The data that support the findings of this study are openly available as detailed in the following. The data source for the observation fields in Figure 1a is included in the AR6 citation: Durack, P. J.