Marine health of the Arabian Gulf: Drivers of pollution and assessment approaches focusing on desalination activities

The Arabian Gulf is one of the most adversely affected marine environments worldwide, which results from combined pollution drivers including climate change, oil and gas activities, and coastal anthropogenic disturbances. Desalination activities are one of the major marine pollution drivers regionally and internationally. Arabian Gulf countries represent a hotspot of desalination activities as they are responsible for nearly 50% of the global desalination capacity. Building desalination plants, up-taking seawater, and discharging untreated brine back into the sea adversely affects the biodiversity of the marine ecosystems. The present review attempted to reveal the potential negative effects of desalination plants on the Gulf’s marine environments. We emphasised different conventional and innovative assessment tools used to assess the health of marine environments and evaluate the damage exerted by desalination activity in the Gulf. Finally, we suggested effective management approaches to tackle the issue including the significance of national regulations and regional cooperation.


Introduction
Marine environments are highly valuable as they host productive ecosystems. These ecosystems provide important goods (e.g. consumable fish, water, and raw materials) and services (e.g. tourism, flood and pollution control, and transportation) for humans (Barbier, 2017). However, marine environments are being continuously altered under the action of different stressors, such as climate change, anthropogenic disturbance, and chemical contamination (Chapman, 2017). Marine pollution and the introduction of exogenous substances have the potential to affect water quality, harm various marine organisms, and affect human health (Beiras, 2018). Therefore, marine ecosystems are currently highly threatened because of marine pollution (Baztan et al., 2016).
The concerns with marine pollution are highly significant because different forms of pollution input to the marine environment lead to changes in the physiochemical and biological properties of the sea, which consequently alter its ecosystems affecting the health and diversity of organisms (Wowk, 2013). A recent study revealed that both biomass-specific primary production and chlorophyll content were significantly reduced due to heavy fuel oil pollution (Lemcke et al., 2019). Furthermore, increasing nutrient concentration in the Gulf of Mexico promoted eutrophication followed by acidification affecting marine ecosystems (Laurent et al., 2017).
Desalination activities are important drivers of marine pollution especially in the Arabian Gulf owing to the high dependency on desalination processes to produce freshwater (Sharifinia et al., 2019). According to Ibrahim and Eltahir (2019), nearly 50% of worldwide seawater desalination is processed in the countries surrounding the Arabian Gulf including Kingdom of Saudi Arabia, Emirates, Kuwait, Qatar, and Bahrain. Desalination processes harm marine habitats owing to the construction of plants, water intake, and brine discharge altering community composition and loss of biodiversity (Sharifinia et al., 2019). Desalination activity in the Arabian Gulf is an emerging issue that must be investigated. Thus, we herein reviewed the most important drivers of pollution in the Arabian Gulf with particular emphasis on desalination plants and their effect on marine health. We also reviewed the different methods that can be used to assess marine health. Finally, several management approaches were proposed to minimise the negative effects of desalination plants in the region.

Drivers of pollution affecting marine health in the Arabian Gulf
The continuous variation in sea temperature and high salinity exert natural stress on the marine environment of the Arabian Gulf (Naser, 2017). According to Joydas et al. (2015), water temperature increases to as high as 36 • C during hot periods of summer and as low as 15 • C in winter, with a salinity that could surpass 43 ppt. Human activities exert an even higher level of stress on the marine environment. In fact, the Arabian Gulf is one of the most anthropogenically affected regions (Halpern et al., 2008). The increasing pollution in the Arabian Gulf poses a great threat to marine habitats and aquatic biodiversity in the region (Sharifinia et al., 2019). The drivers of pollution in the Arabian Gulf are diverse, but climate change, oil and gas industry, coastal anthropogenic disturbance, and desalination plants are considered to be the most important sources of pollution (Fig. 1).

Climate change
Climate change is a global issue threatening aquatic ecosystems (Henson et al., 2017). It is an undeniable phenomenon altering the biogeochemistry of seas, which is triggered by increased temperature (Gattuso et al., 2015). Ocean warming is coupled with increasing stratification of the sea, which prevents nutrients from moving to the upper photic layers to support photosynthetic microorganisms (Steinacher et al., 2010). Climate change also induces the solubility of CO 2 into the seas due to its high concentration in the air (Henson et al., 2017). This results in decreasing the pH of oceans and negatively affecting calcareous organisms (Doney et al., 2012). Besides, oxygen solubility in the water will decrease as water temperature increases affecting aquatic organisms (Vaquer-Sunyer and Duarte, 2008).
Marine organisms are able to cope with such changes to some extent through physiological and phenological adaptation, as well as shifting population distribution and dynamics (Wabnitz et al., 2018). The result of these changes is reflected in species richness (Jones and Cheung, 2015), and also affects the structure of marine communities (MacNeil et al., 2010). According to Wabnitz et al. (2018), climate change is increasing the pressure on the Arabian Gulf, resulting in the survival of certain organisms that tolerate extreme temperature and high salinity. Despite the high adaptability of some organisms, climate change still exerts negative effects on them. For instance, increased sea temperature in the Arabian Gulf resulting from climate change caused enormous bleaching followed by death in corals in 1996 and 1998 (Naser, 2014).

Desalination plant discharges
Arabian Gulf countries are characterised by an arid climate with a scarcity of freshwater (Elasha, 2010). Supplying the needed freshwater to the region is achieved by depending on desalination plants. Even though desalination processes are beneficial as they provide an important source of freshwater, they have a significant negative environmental effect on the marine chemistry and health with increasing the salinity, temperature, heavy metal concentration which ultimately alter the marine biodiversity (Roberts et al., 2010). With increasing populations and economies, desalination plants, as well as combined water and power production plants, are increasing in coastal areas of the Gulf. In fact, coastal infrastructure for water, energy, and food supply is becoming increasingly coupled with some mega plants supplying major cities in the region, thus posing several risks to coastal populations and marine ecosystems (Al-Saidi and Saliba, 2019). This is an emerging issue threatening the health of the Arabian Gulf.

Oil and gas activities
Arabian Gulf countries are the largest producers of oil and gas globally, and it is estimated that 65% of proven crude oil reserves belong to Arabian Gulf littoral countries (Modarress et al., 2016). By 2035, oil exports from the region are expected to exceed 43 million barrels per day (Modarress et al., 2016). The development of oil and gas industry has caused destruction in the Arabian Gulf's marine environment. This is especially true when considering the activities associated with such developments, such as the construction of platforms, extraction of oil and gas, refining, and transportation.
As many spills and oil-related accidents go undocumented, it is difficult to quantify the effects of these events (Meshkati et al., 2016). Nonetheless, one of the largest oil spills happened in 1991 when 10 million barrels of oil were spilled during the Gulf War resulting in a long oil slick covering the Kuwaiti and Saudi coasts (Alhanaee et al., 2017). Oil pollution is believed to have an important effect on biota and biodiversity (Azevedo-Santos et al., 2016). Acute oil pollution causes mass mortality of organisms (McGenity et al., 2012) as well as substantial decreases in species richness (De La Huz et al., 2005). In their review, Pashaei et al. (2015) indicated that oil pollution and spills in the Arabian Gulf severely damaged mangrove forests, killing more than 500 sea turtles, decreasing 25% of shrimp fisheries, and polluting sediments.

Coastal anthropogenic disturbances
Coastal anthropogenic disturbances are a set of all human activities that induce damage to the aquatic habitats. They include coastal modification, dredging, land reclamation, and other activities that result in water pollution and habitat destruction (Naser, 2017). Arabian Gulf countries have experienced substantial development in economic and industrial sectors, which have resulted in considerable modification of the Gulf coastline to accommodate ports, artificial islands, marinas, coastal hotels, and even maritime cities, all of which increased marine pollution (Naser, 2015). For instance, high concentrations of heavy metals and hydrocarbons were reported in areas of active shipping in the Arabian Gulf (de Mora et al., 2010). Another study done in Saudi Arabian coasts of the Arabian Gulf revealed that the concentrations of heavy metals including zinc, copper, chromium, and lead were enriched due to anthropogenic activities (Almasoud et al., 2015). Also, studying the PAHs concentrations in Qatari coastal sediments revealed moderate to high pollution in Al-wakrah port harbour (Soliman et al., 2014).
The introduction of such toxicants into coasts often promotes losing richness and biodiversity of marine species (Johnston and Roberts, 2009). Abd El-Wahab and Al-Rashed (2014) indicated that coastal habitats of Kuwait are negatively affected by human activities as both species composition and diversity of plants were considerably altered over the last five decades. In addition, based on Loughland et al. (2012), around 90% of native saltmarshes in the Arabian Gulf were lost due to coastal development and urbanisation. Baby et al. (2014) suggested that the carrying capacities of Kuwaiti's coastal habitats are decreasing due to urbanisation and industrialisation.

Desalination plants in the Arabian Gulf
Freshwater is a finite resource. Owing to the rapid increase in human population, unsustainable consumption, and the global changes in climatic conditions, water scarcity threatens many parts of the world (Odhiambo, 2017). The Arabian Peninsula has some of the scarcest freshwater resources worldwide (Elasha, 2010). This increased demand compared to the limited supply of freshwater drove the Arabian Gulf countries and especially the Gulf Cooperation Council (GCC) to rely on seawater desalination. It is estimated that the Arabian Gulf countries have a desalination capacity of 11 million m 3 per day. The region accommodates 213 actively operating desalination plants, and 51 others are expected to be commissioned in the near future (Sharifinia et al., 2019). Saudi Arabia (45%) and the United Arab Emirates (22%) are the two biggest contributors to desalination activities in the region.

Types of desalination plants
Desalination plants differ based on the separation technique used. There are two main classifications of desalination plants, namely thermal-based or membrane-based separation (Table 1). Globally, the most widely used desalination systems are based on reverse osmosis (RO) followed by multi-effect distillation (Miller et al., 2015). RO alone accounts for more than 60% of globally produced desalinated water. There are only three types of desalination plants present in the Arabian Gulf. These are RO, multi-effect distillation, and multi-stage flash. Multistage flash distillation accounts for 81% of water desalination, multi- -Based on evaporating the incoming water using heat that comes from vapor compression.
-Mechanical compressors are often utilised for heat generation.
-Usually have small capacities and they could be coupled with multi-effect distillation (MED). (Krishna, 2004) Multi-Stage Flash distillation (MSF) -The influent water is heated to 120 • C under high pressure. Heated water passes through successive flash units.
-Each unit has a lower pressure compared to the one before allowing the hot water to evaporate as it flows through the flash units.
-The evaporated portion is cooled and condensed on heat exchanger tubes.
-Condensed water is collected as freshwater while the brine leaves the flash units to be discharged. Membrane-based desalination Reverse osmosis (RO) -RO is the most commonly used membrane-based desalination processes.
-It is a filtration process in which water is forced into a membrane that separates freshwater from brine.
-In definition, osmosis is the movement of water through a semipermeable membrane from a solution having fewer salts to a solution having more salts.
-In reverse osmosis, pressure is applied to reverse the direction of water flow, which produces freshwater and concentrated brine. (Miller et al., 2015) (Krishna, 2004) Electrodialysis (ED) -Based on the fact that dissolved salts are ions bearing either a negative or a positive charge.
-ED utilises selective membranes that allows positively charged ions or negatively charged ions to pass through upon receiving electric currents.
-Anion selective membranes and cation selective membranes are arranged alternatively in the desalination plant to separate all salts from freshwater. (Krishna, 2004) Electrodialysis Reversal (EDR) -EDR process was launched after the ED.
-It is very similar to EDR in terms of operation principle.
-The main difference is that the polarity of the direct current is reversed several times an hour.
-This attracts ions to the opposite direction of the membrane. Afterwards, the freshwater is collected. Both ED and EDR are mostly used for brackish water rather than seawater. (Buros, 2000) (Valero and Arbós, 2010) (Krishna, 2004) effect distillation accounts for 13%, and RO accounts for only 6% (Sharifinia et al., 2019). Thermal desalination is the predominant technology due to the abundance of fossil fuels in GCC countries, making fossil-fuel-dependent powerplants the most economically attractive method for generating energy to drive desalination processes (Dawoud and Al Mulla, 2012).

Effect of desalination on marine environments
The effect of desalination activities on the marine environment has not been widely studied. In fact, according to Kress et al. (2020), the majority of the existing publications present predicted, potential impacts which are not based on observed or experimental data. In spite of that, based on the existing data, the effects could be categorised into two categories based on the intake of seawater and discharge of brine. Generally, desalination activities alter the structure and function of marine ecosystems, which is visualized by the affected marine communities and changes in the trophic interactions (Grossowicz et al., 2020).

Intake of seawater
Desalination plants rely on water from different sources, e.g. cooling water used in power plants, aquifers, ground water, and most commonly, open seas. When desalination plants are being constructed, pipes are installed to transport water from the sea. This step disrupts the seabed causing resuspension of sedimented particles, including pollutants (Dawoud and Al Mulla, 2012). Disturbance and alteration of the seabed leads to habitat destruction, death of marine species, release of toxic pollutants from sediments, and increase in water turbidity.
Once plants are operating, massive volumes of water are pumped into the plants directly from the sea. These volumes are estimated to be double the amounts being produced (Kress et al., 2018). Along with water, many organisms are taken into the system that either get impinged (crash into the screens of the intake pipes) or entrained (travel with water reaching the plant) (Dawoud and Al Mulla, 2012). Entrainment and impingement result in severe injury and death of marine organisms (National Research Council, 2008). According to Missimer and Maliva (2018), assessing the effect on the marine environment resulting from entrainment and impingement is difficult. Parameters such as screen mesh size, pipe size, and volume of water intake, should be considered when designing the plant to reduce entrainment and impingement (National Research Council, 2008). Subsurface water intake reflects another potential solution (Missimer et al., 2013).

Discharge of the brine
Desalination processes result in the production of brine, which is a waste fluid characterised by high salinity and dissolved minerals (Danoun, 2007). The fate of the brine is usually disposal into the sea, considering it as one of the least costly disposal approaches (Fernández-Torquemada et al., 2019). The desalination process goes through various phases; thus, the produced brine contains different chemicals and agents. The first important aspect to consider is the high salinity of the brine, which is at least 1.6-2 times higher than that of seawater (35 g/L average) (Panagopoulos et al., 2019). Furthermore, depending on the technology used (i.e. thermal technologies) brine can exceed seawater ambient temperature by 1.37 to 1.82 times (Missimer and Maliva, 2018). The brine can also contain different chemicals, such as chlorine, cationic and anionic coagulants, acids, anti-scalants, heavy metals, and anti-foaming agents that are added during the desalination process (Alameddine and El-Fadel, 2007;Frank et al., 2019). Once discharged into the sea, these chemicals are considered to be toxic pollutants. Furthermore, the brine is highly alkaline as a result of calcium carbonates and sulfates (Danoun, 2007). The discharge of brine with these characteristics into the sea leads to significant changes in the physiochemical and biological parameters of the sea which ultimately affect marine life.
High temperature and salinity inhibit the growth of aquatic organisms (Wiltshire et al., 2010). Salinity elevations affect marine organisms such as planktons, microbes, and benthic species . In addition studies have shown that increasing the salinity of aquatic environments slightly above ambient conditions disrupts the osmotic regulatory abilities of some marine organisms resulting in dehydration and consequently death (Al-Shammari and Ali, 2018; Matsumoto and Martin, 2008). In addition, increasing the temperature of seawater increases the toxicity of some chemicals and metals, which adversely affects aquatic life (Uddin, 2014). Research has focused on seagrasses given the importance of seagrass habitats, which comprise diverse organisms and are sensitive to fluctuations in environmental conditions (Kress et al., 2018). Table 2 summarises research on the effect of increased salinity and temperature as a result of brine discharge into the sea.
Brine disposal also causes hypoxia resulting from decreasing concentrations of dissolved oxygen, which affect all marine organisms (Ahmed and Anwar, 2012). Increasing salinity is inversely proportional to the concentration of dissolved oxygen in the sea (Krayer et al., 2017). In addition, increasing water temperature through input of hot brine decreases oxygen solubility (Ahlgren et al., 2017). Oxygen-depleted water bodies experience mass mortality of mussels, bivalves, and fish, and also disruption to coral reef functionality, invasion of opportunistic jellyfishes, and loss of biodiversity (Isensee and Valdes, 2018). The significance of brine discharge is arguably high in the Arabian Gulf considering that the Gulf is shallow and semi-enclosed with weak water circulation and limited freshwater input (Uddin et al., 2011). Such conditions accommodate adapted native species, which are resistant to fluctuations in the physiochemical parameters of the sea. Therefore, the extensive desalination activity in the region greatly threatens sensitive species, possibly leading to their extinction.

Methods for assessing marine ecosystem response to stressors
Increased population and associated activities are exerting huge amounts of pressure on the marine environment (Halpern et al., 2015). Despite realising the importance and fragility of marine resources, we continue to exploit, destroy, and pollute the oceans, which leads to losses in functionality and biodiversity of aquatic ecosystems (Claudet and Fraschetti, 2010;McCauley et al., 2015). As a counter measure, many laws and regulations have been implemented globally aiming to protect the marine environment and conserve its ecosystems. Such regulations rely on our ability to assess marine health using different interconnected tools (Boyes and Elliott, 2014).
Assessing the health of the marine environment requires combining different parameters to reach a realistic conclusion. It is essential to adopt different assessment approaches including incorporating physical, chemical, and biological parameters. Borja et al. (2016) insisted on the significance of including biotic and abiotic factors, as well as human and social intervention when assessing environmental status. To develop a complete perspective of diversity indicators, biotic components should be incorporated into ecosystem assessments at various levels, including genus, species, population, and community levels (Haase et al., 2018). Considering that the assessment of each level serves a different objective, combining more than one would increase the objectivity of judgments of ecosystem health. In this context, various tools are implemented to assess marine health, most of which are oriented towards studying the diversity of marine organisms. There are two main categories of assessment tool comprising conventional methods and innovative recently developed strategies (Fig. 2).

Conventional tools for the assessment of marine health
Conventional tools are those that were historically used to monitor environmental health. They include several methods and approaches that were extensively used until the beginning of the 21st century.
Thereafter, such approaches were amended by innovative methods owing to the development of methods to avoid the limitations of conventional techniques.

Biodiversity and abundance of faunal communities
Diversity is one of the most important parameters for examining the effect of human activities on marine ecosystems. Many diversity indices have been developed to interpret the data collected and relate these to assessments (Yoccoz et al., 2001). Indices are mostly based on the relative abundance of each species (Yoccoz et al., 2001). Shannon-Wiener and Simpson are the most widely used biodiversity indices (Mendes et al., 2008) even though they are highly influenced by sample size, making them biased and, thus, less reliable (Hewitt et al., 2005).
Determining the biomass and abundance of faunal species has been widely used to monitor the marine environment. According to Pagola-Carte and Saiz-Salinas (2001), studying benthos can reveal important information about environmental conditions. Analysing the biomass and abundance of benthic species depends on primitive and simple protocols which involve divers collecting all benthos that fall within a quadrant of known area and quantifying and identifying the benthos (e.g. Barnes and Brockington, 2003). Strong et al. (2015) considered that these structural indicator studies are widely used for environmental monitoring because they are well established and inexpensive. However, indices are not yet highly informative regarding the functionality of the ecosystems. Similarly, Bremner et al. (2003) indicated that such studies can provide information on human effects on marine ecosystems at the community level, even though, they poorly address ecological functioning.

Ecological indicators
Ecological indicators are species that are either highly sensitivity or tolerant to changes in an ecosystem. Studying the richness and abundance of indicator species provides rich information on marine health and response to pollution (Aguirre and Tabor, 2004; Parmar et al.,

Table 2
The effect of increased salinity and temperature as a result of brine discharge.  Fig. 2. Different approaches adopted for the assessment of environmental effect of desalination on marine health. 2016). On the one hand, sensitive species are usually dominant in marine systems and a reduction in their abundance indicates an altered ecosystem. On the other hand, tolerant species are opportunistic and highly resistant to environmental changes. Increase in the number of tolerant species is an indication of an altered ecosystem (Simboura and Zenetos, 2002). Fish are informative indicators as they are mobile, long lived, and they are present in all aquatic environments (Whitfield and Elliott, 2002). Indicator species are used to study the effect of human activities on marine ecosystems. For example, phytoplankton are informative bioindicators of eutrophicated waterbodies, where they tend to grow rapidly (Singh et al., 2013). As indicated by Anttila et al. (2018), blooms of cyanobacteria in the Baltic Sea always indicate anthropogenic increase in the nutrients inputs leading to eutrophication. Furthermore, as described by Hosmani (2014), zooplankton are sensitive to changes in the physiochemical conditions of water (e.g. chemical composition, dissolved oxygen, pH, and temperature), which is why they are used as bioindicators. Algae were also used as indicators of organic pollution in lakes and freshwater bodies (Hosmani, 2013). Even though relying on indicator species for the assessment of environmental health is relatively easy and cost-effective, there are many shortcomings associated with these methods (Siddig et al., 2016). Firstly, considering a single population would not sufficiently represent the effect on the ecosystem as a whole. Secondly, using this approach neglects the biological interactions and its influence on the population of the indicator species. Finally, environmental factors other than the studied pollution (e.g. global warming) might influence the indicator species.

Biomarkers
In definition, biomarkers are changes that occur biologically, biochemically, or physiologically to organisms as a result of exposure to xenobiotic compounds (Hahn, 2002). Biomarkers have been incorporated in the ecotoxicological tests, where the existence of pollutants in a certain environment could be measured at the molecular level of affected organisms . This method enables the identification of pollutant toxicity in exposed organisms (Galloway et al., 2002). It is necessary to develop standards and define norms of biomarkers for every organism (Viarengo et al., 2000) to be able to compare experimental data with reference values. It is more advantageous to study multi-biomarkers to understand the effect of stressors on organisms at the molecular level (Downs et al., 2002). This technique is not routinely used due to the difficulties associated with interpreting the acquired data because the response of biomarkers is not completely understood at different biological levels (Brown et al., 2004). Furthermore, biomarkers tend to be influenced by different factors other than pollution, e.g. organism age, salinity, and water temperature (Brown et al., 2004). Therefore, it can be difficult and complicated to link the pollution with its effect on biomarkers (Viarengo et al., 2000). Additionally, there are difficulties associated with selecting the biomarker that is representative of the changes occurring.

Innovative and novel methods for the assessment of marine health
Marine monitoring has always been challenging owing to the difficulties associated with providing impartial data, knowing that marine ecosystems are highly complex structurally and functionally (de Jonge et al., 2006). Traditional methods of monitoring are widely used as they are cost effective, widely accepted, and well established, although they still have many limitations (Bourlat et al., 2013;Strong et al., 2015). Conventional monitoring strategies are usually restricted by the sampling site, and are usually limited by the fact that the monitoring focuses on specific organisms that are easily monitored, which leads to inaccurate and biased results (Bourlat et al., 2013). When using conventional methods, marine environment health is assessed by targeting a certain taxonomic group during a certain life stage. Other interactions are mostly neglected, thus causing a lack in understanding of ecosystem interactions, and consequently, a poor understanding of the effect of the studied stressor on the marine ecosystem as a whole. Consequently, innovative strategies for monitoring and assessing the health of marine environments had to be developed. There have been different marine assessment tools emerging recently, and they are very promising as they overcome the shortcomings of conventional approaches. The main methods that fall under this category are acoustic devices, remote sensing, and genomics .

Acoustic devices
An advanced method for monitoring marine ecosystems involves utilising acoustic devices to detect changes caused by human activity. Marine mammals are known for their acoustic specialisation where they rely on sound waves to communicate and navigate (Sousa-Lima et al., 2013). Owing to technological developments, we now have devices that can record underwater sounds produced by marine mammals. Following these developments, passive acoustic monitoring (PAM) devices were developed as non-invasive tools for better understanding marine mammals and the sounds they produce (Kalan et al., 2015). PAM is used to monitor animals by utilising remote acoustic tools such as hydrophones, stereo-phones, microphone arrays, and other auto-recording technologies (Marques et al., 2013). Species richness and community composition could be determined using PAM (Blumstein et al., 2011). PAM has proven to be highly beneficial when it comes to marine mammals because they are especially difficult to monitor optically because light cannot penetrate oceans for long distances. PAM systems can detect and classify marine mammals (Bittle and Duncan, 2013), which is why they are useful for studying the effect of human activities on marine health. Even though they are highly useful, PAM devices comprise continuous real-time monitoring systems that are associated with the long-term recording of vast amounts of data (Lammers et al., 2008). This is considered to be a major limitation as it requires extensive inspection and analysis of these data, which is usually unfeasible (Swiston and Mennill, 2009).
As an alternative, underwater fixed autonomous, sound recorders were developed in the early 1990s which addressed the limitations of PAM. These new systems are characterised by lower costs, and they do not require experts for continuous sound monitoring (Sousa-Lima et al., 2013;Wiggins et al., 2012). Autonomous sound recorders record soundwaves and store the collected data within the device without being connected to reception stations (Wiggins et al., 2012). They do not require running and monitoring in person as they are installed on buoys or fixed to the sea floor where they record continuously. They could also be incorporated into autonomous underwater vehicles like ocean gliders (Fucile et al., 2006). After a defined period of time, the devices must be retrieved to analyse the acquired data (Sousa-Lima et al., 2013).
Using autonomous sound recorders could be highly advantageous for remote areas (e.g. polar regions) or when access to the location is difficult (e.g. deep sea or harsh weather) (Širović et al., 2009;Soldevilla et al., 2010). sing these acoustic devices could be an efficient means to indirectly measure the biodiversity of acoustic mammals to assess ecosystem health (Sueur et al., 2008). Acoustic data could be analysed and interpreted through a variety of ways serving different purposes. For example, biodiversity could be assessed by quantifying spectral and temporal entropy H (Sueur et al., 2008). Parks et al. (2014) discovered that using noise compensated entropy (H N ) was the most representative index that reflects biological patterns and diversity in the marine environment. Despite the advantages of acoustic assessment, there remain issues concerning the interpretation of complex data. Besides, it is an indirect measure of biodiversity which is not always able to provide representative results.

Remote sensing
Remote sensing is an innovative approach recently used to assess marine health. Remote sensing, including optical, thermal, and radar sensors, provides new prospects for studying species, habitat distribution, and biodiversity (Pettorelli et al., 2014;Turner et al., 2003). This approach was previously used to monitor terrestrial ecosystems (Gross et al., 2009). However, recently, this technology was developed to include monitoring aquatic environments and was successfully implied for studying, monitoring, and managing mangrove (Kuenzer et al., 2011), coral reef (Hamel and Andréfouët, 2010), and even seagrass (Dekker et al., 2006) ecosystems.
According to van der Wal and Herman (2007), radar imaging is efficient at determining the composition of saltmarshes and intertidal habitats. Remote sensing proved to be cost effective for in-situ sampling, which provides data based on spatial and temporal screening of marine communities enabling scientists to understand the effect of human activities on the dynamics of marine communities (Rivas et al., 2006). According to Blondeau-Patissier et al. (2004), the development of remote sensing enabled spanning temporal and spatial recordings of sites, which overcomes the disadvantages of conventional in-situ marine monitoring techniques.

Assessing productivity.
Owing to the large size of the seas and oceans, it is difficult to rely only on buoys and ships for monitoring purposes. Remote sensing is more informative, especially when it comes to assessing primary productivity. Since chlorophyll a is the main pigment for photosynthesis, its concentration could reflect primary productivity. The concentration of the pigment is determined using different remote sensing techniques, which operate in the visible region of the light spectrum (Klemas, 2010). Multispectral and hyperspectral imagers are used to measure chlorophyll concentration and thereby assess primary productivity using sensors like SeaWiFS and MODIS (Oliver et al., 2004). Chlorophyll is measured based on atmospheric spectral radiance, which is used to derive the spectral radiance of ocean surfaces (Bagheri et al., 2002). Derived surface radiance is then used to calculate the reflectance which is important for chlorophyll identification and measurement (Philpot, 2007). According to Klemas (2010), obtaining accurate and calibrated data requires coupling the remote sensing data with data collected using ocean gliders, ships, and buoys.

Assessing the health of coral reefs.
Remote sensing has been efficiently used to monitor coral reefs. Since coral reefs are highly fragile environments, it is important to monitor the effects of human activities on such ecosystems (Klemas, 2010). Coral studies using remote sensing could be direct in which data are related to the reef itself (Wabnitz et al., 2010), or indirect in which data represent the environmental conditions of the reef. Direct measurements include the location of reefs, patchiness, cover, and diversity of the habitat (Hamel and Andréfouët, 2010). Conversely, indirect measurements refer to temperature, turbidity, chlorophyll concentration, and organic matter concentration of the oceans, and also wind, rain, and cloud cover in the atmosphere (Hamel and Andréfouët, 2010). Coral reefs and submerged vegetation can be mapped by both hyperspectral and multispectral imagers (Akins et al., 2009).
Even though remote sensing represents a promising technique, there are still limitations that prevent such techniques from operating at full efficiency. Firstly, remote sensing observations and data are restricted to clear days because the presence of clouds restricts the collection of data especially in tropical and high-latitude areas (Peters et al., 2005). Furthermore, measuring the concentration of chlorophyll a to estimate the biomass of phytoplankton for productivity assessment is based on conversion factors which could be misleading (Rivas et al., 2006).

Genomics
Genomic assessment of marine environments provides information on marine ecosystems at the cellular and microbial levels. This type of assessment provides information about micro-communities, their interactions, and the involved metabolic pathways, which allows a comprehensive evaluation of the functionality of an ecosystem (Bourlat et al., 2013), and provides a representation of the ecosystem's current and future response to stressors in terms of the effect on organisms and their interactions. Consequently, this will represent changes in populations and communities . Genomics techniques are considered to be an emerging tool and have shown promise for environmental monitoring. They essentially provide cost-effective and reliable measurements, which is why they are expected to substitute traditional methods (Bourlat et al., 2013).
As genomic approaches are becoming more widely applied owing to the technological advancement of sequencing tools (Mardis, 2008), genetic information for different species and habitats has become widely available in databases (Bik et al., 2012;Hajibabaei et al., 2011), which has facilitated the analysis of genetic data acquired from different ecosystems, including the marine environment. Following these developments, the methodologies of molecular analysis are constantly being refined, developing novel methods for different purposes (Leese et al., 2016).

4.2.
3.1. DNA barcoding and meta barcoding. DNA barcoding involves sequencing and analysing standard short fragments of DNA known as DNA barcodes to identify the species of an unknown biological sample (Hebert et al., 2003). This type of genomic analysis is simple and does not require taxonomic experts as the sequenced barcodes are simply compared with other barcode sequences in a database to identify the species of the unknown sample. This technique is advantageous as it could be applied not only to biological specimens and tissues but also environmental samples for identifying all taxa in an area by metabarcoding (Bourlat et al., 2013). With metabarcoding, it is possible to analyse DNA present in water, biofilms, and even sediment samples, which is referred to as environmental DNA (eDNA) (Leese et al., 2016).
Using this approach allows the assessment of ecosystem's biodiversity through monitoring multiple communities at once (Zhang et al., 2020). DNA barcoding and metabarcoding require a DNA barcode library that contains information about different species. Constructing such reference libraries requires expert taxonomists and extensive effort. In their review, Taylor and Harris (2012) mentioned that the main resources for DNA barcodes are Consortium for the Barcode of Life (CBOL), International Barcode of Life (iBOL), and the Barcode of Life Data System (BOLD).

Metagenomics.
The conventional study of microbial species is restricted to culturable species, which are in the minority in the microbial world. Metagenomics offer an innovative approach to study culturing-recalcitrant microbial fractions through the DNA sequencing of any environmental sample (Culligan and Sleator, 2016). This approach is promising especially with the development of nextgeneration sequencing (DiBattista et al., 2020;Schuster, 2008). It allows rapid environmental monitoring through relative abundance estimation (Günther et al., 2018). There are mainly two types of metagenomic method involving either 16S rRNA amplicons or whole genome shotgun sequencing.
• 16S rRNA amplicons Combining the phylogenetic diversity approach with metagenomics would simplify assessments of the biodiversity of environmental samples using marker genes. According to McDonald et al. (2013), determining phylogenetic diversity depends on studying marker genes such as 16S rRNA to know the similarity between microbial communities. As with DNA barcoding, metagenomics relies on reference libraries of genetic data of species. 16S rRNA techniques have been recently developed following the advancements in next-generation sequencing strategies that utilise 16S rRNA primers, such as the Illumina sequencer and 454 pyro-sequencer (Shah et al., 2011). Despite its efficiency in studying microbial diversity, the results provided by 16S rRNA metagenomic studies might not always be reliable since DNA extraction and PCR amplification procedures can produce biased data (Brooks et al., 2015).
• Whole-genome shotgun This type of metagenomic technique sequences eDNA using random primers which results in overlapping genomic sequences (Ranjan et al., 2016). Such studies provide information about the genomic and metabolic characteristics of environmental samples (Kalyuzhnaya et al., 2008) Whole-genome shotgun sequencing could be favoured over amplicon sequencing because it is more objective and less biased. However, it is not widely used because it is expensive and requires analysis of vast amounts of data (Luo et al., 2014;Sims et al., 2014). It is important to mention that whole-genome shotgun sequencing has other limitations, such as lacking the ability to detect rare species (Kalyuzhnaya et al., 2008).

Microarrays.
Microarrays are chips that contain a collection of labelled DNA probes, each probe representing a different species (Bourlat et al., 2013). DNA from environmental samples hybridises with a probe forming a complex, which will fluoresce upon subjecting it to UV radiation. Such techniques represent an innovative strategy for assessing environmental health (He et al., 2007), and allow for the detection of harmful microbes, or can infer the absence of locally dominant species in environmental samples, which could occur as a result of stressors. For example, studies have used microarrays as efficient and rapid ways for identifying toxic algal species that are responsible for harmful algal blooms (Bricker et al., 2008;Doucette et al., 2009).

Quantitative real-time PCR.
This method is based on quantifying a certain gene sequence belonging to a specific organism. DNA is quantified by comparing it to known values from standards curves and then it is correlated with the number of individuals of that species (Bourlat et al., 2013). This is a simple assessment tool for marine health, where knowing the abundance of a species allows us to determine whether that species is being affected by pollution. This technique can also be used to assess the genetic diversity in the affected environment (Smith and Osborn, 2009). Real-time PCR assays have high sensitivity and quantification power, making it a reliable marine monitoring tool (LeBlanc et al., 2020) However, even though this method is helpful for assessing marine health, it can only be applied to unicellular organisms with a known number of copies of the gene being quantified (Bourlat et al., 2013).

Transcriptomics.
Transcriptomics is the study of gene expression. In the context of environmental monitoring, this study provides a comprehensive view of organisms' responses to stress at the molecular level (Devens et al., 2020). Gene expression can be studied through different techniques, such as using real-time PCR to quantify the concentration of RNA or through RNAseq (Bourlat et al., 2013).
In the Arabian Gulf, out of all of the assessment tools reviewed, genomics seems to be the most promising approach because it overcomes many of the limitations associated with the conventional assessment tools, complexity of interpreting acoustic data, and restrictions of remote sensing. Genomics approaches will provide an overall evaluation of marine ecosystems including their functionality. It will reveal the current status of the Arabian Gulf and possible future responses through DNA analysis.

Management approaches for future coastal development
Since marine environments comprise complex interacting ecosystems that provide extensive goods and services (Barbier, 2017), it is important to protect them. The extensive desalination activity in the Arabian Gulf is a potential threat to marine health. The effects of desalination plants can be limited by applying effective mitigation measures. Some of the important management approaches are presented in this section.

Optimisation of plant design
The negative effect of the desalination process could be reduced by optimising the design of the plant. Using membrane-based desalination could be less destructive to marine habitats since thermal-based desalination produces hot brine that adversely affects marine life when discharged. Besides, the inlet of the intake pipes should be located in places with low species abundance, avoiding productive areas to reduce the negative effect. Furthermore, the entrainment and impingement of organisms could be minimised by optimising the velocity of water flowing through the intake pipes and by optimising the mesh size of the screens in the intake pipes (Sharifinia et al., 2019). In addition, developing and using environmentally friendly desalination processes should be emphasised. For instance, solar-based desalination processes are being developed to increase their efficiency and applicability. Palenzuela et al. (2015) proposed coupling concentrated solar power plants with desalination plants in the Arabian Gulf to produce electric power and freshwater.

Treatment of brine before discharge
To minimise the negative effect of brine discharge, it is possible to treat the brine to remove hazardous chemicals used during the desalination process. Chlorine is an example of a hazardous chemical used during desalination since it has a biocide activity. Lattemann and Höpner (2008) indicated that chlorine could be removed using sodium bisulfite for RO plant effluents, and hydrogen peroxide for thermal plant effluents. Source control is even more effective where hazardous chemicals are not used during the desalination process. Such chemicals have substitutes that are less environmentally destructive. In addition, many studies have focused on treating produced brine by enhancing water recovery. As reviewed by Panagopoulos et al. (2019), brine treatment technologies could be membrane based, thermal based, or zero liquid discharge based. These technologies work principally on producing pure water and compressed solids, which reduces brine volume.

National regulation
National monitoring and regulation are required for assessing the health of ecosystems and setting guidelines and rules for plant design and discharge parameters. For example, in response to the increasing pollution in the Arabian Gulf region, Kuwait established the Kuwait Environment Public Authority (KEPA). KEPA is an independent organisation responsible for maintaining the health of the environment and actively participates in enforcing legislation and setting standards. As a regulatory measure, KEPA monitors the quality of Kuwait's territorial water through continuous collection of data from 13 different stations. The quality of water is assessed based on different parameters to make sure it meets the local standards (Al-Mutairi et al., 2014). To control the increase in salinity of the sea resulting from desalination activities, KEPA set a salinity limit of 42 ppt (Uddin et al., 2011). This is an effective initiative for controlling the direct dumping of brine into the sea.

Regional cooperation
Since the Gulf waters are shared among several littoral countries, a regional approach to the control of marine pollution is indispensable. In this sense, Gulf littoral countries adopted a legal instrument in 1979, namely the Kuwait Regional Convention for Cooperation on the Protection of the Marine Environment from Pollution (short Kuwait Convention). Based on this convention, the Regional Organisation for the Protection of Marine Environment (ROPME) was established in 1979 as an intergovernmental organisation encompassing all Gulf littoral countries. A cornerstone of ROPME policy includes an action plan (the Kuwait Action Plan) for monitoring and assessing the health of the Gulf. Furthermore, several additional protocols to the Kuwait Convention have been signed over the years, targeting pollution control from sources such as oil, exploration activities, land-based activities, shipping, and the disposal of hazardous materials. An additional protocol for biodiversity and protected areas has been discussed since the early 2000s but has not yet been implemented.
ROMPE has been important for promoting cooperation on marine pollution and initiating joint action, e.g. through ROPME Marine Emergency Mutual Aid Center (ROMPE/MEMAC) in Bahrain focusing on coordinating the Gulf's response to oil spills. The issue of brine discharge is part of the additional protocol on pollution from land-based sources adopted in the 1990s, which stipulates that regional regulations, programs, timetables, and measures need to be implemented to address pollution. However, regional action through ROPME has been limited to monitoring and analyses supported by international organisations. Integrating the increasingly important issue of brine discharge into future cooperative frameworks is an important step to effectively address this issue at the Gulf-wide scale. This also means shifting focus from analysing pollution and promoting protection to adopting more comprehensive measures including joint monitoring and regulation (Van Lavieren and Klaus, 2013). Further broader approaches that incorporate biodiversity protection and ecosystem management have been part of the original mandate of ROPME but are not yet reflected in the practice of regional cooperation (Hamza and Munawar, 2009;Khan, 2008).

Final remarks
The Arabian Gulf is considered one of the most anthropogenically affected seas. It is facing different types of stressors inducing marine pollution, which affect the health of the marine environment. It is undeniable that climate change, oil and gas activities, and coastal reclamation are major contributors to this problem. However, recent desalination has been a debatable contributor to marine pollution. Owing to the scarcity of freshwater resources, Arabian Gulf countries rely on desalinating seawater to produce freshwater, which is used for drinking, agriculture, and other purposes. For the longest time, desalination was thought to be a solution for the water scarcity issue in the region. However, recently, negative environmental effects of extensive desalination activity have been revealed. The Arabian Gulf mostly uses thermal-based desalination processes, which result in discharging huge amounts of hot, salt-concentrated brine directly into the sea. Discharge of the brine became a forefront of policy debates after it was neglected, especially in the Gulf area where desalination is extensive, and the natural environmental conditions are harsh.
Considerable effects of brine discharge have been reported worldwide necessitating the adoption of efficient monitoring systems in the Arabian Gulf. Monitoring should develop beyond simple measurements into more integrative, adaptive, and multivariate technologies. For that, genomic monitoring is the most efficient in terms of ease, applicability, and objective representation. Since the Arabian Gulf is considered a hot spot of desalination activities, it is important to implement management approaches at different levels to reduce the negative effect on marine ecosystems. Firstly, optimisation of plant design is a promising approach where many engineering parameters should be considered. Secondly, instead of direct discharge, brine should be treated pre-discharge. Finally, following in the footsteps of KEPA, laws and regulations for desalination activities should be implemented to prevent further damage to the Gulf's environment. Monitoring tools and assessment of the desalination activities should be incorporated into the regulations and policies to prevent future destruction of marine ecosystems.

Funding
This report was supported by the NPRP award [NPRP11S-0110-180248] from the Qatar National Research Fund (a member of The Qatar Foundation).

Declaration of competing interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this review paper.