Abstract
Coastal areas are characterized by a high population and a wide range of industrial and agricultural activities, which puts them under high pressure and continuous pollution from anthropogenic activities. This research focuses on the application of HYDROLAB HL7 multiparameter sonde equipped with smart sensors for the measurement of physical–chemical parameters in marine waters in the Durrës Bay. The sonde is part of a transnational repository network that receives, stores, and analyzes data about seawater quality, serving as an early warning system for preventing the diffusion of marine pollution. This sophisticated instrument can thrive in demanding environmental conditions for long-term continuous monitoring. It maximizes deployment lifespan, and provides traceable data for high-quality, reliable monitoring of vital changes in water quality. Low variability on the measured parameters indicates a stable status in the water quality of the Durrës site. Time series revealed small seasonal variations on all parameters, except turbidity and water temperature. Total dissolved solids, salinity, and electrical conductivity revealed similar temporal trends over the monitoring period by indicating strong relationships between them. The obtained data for the physical–chemical parameters in this study align with the recommended values. Ensuring water quality in the Durrës Bay requires advanced monitoring, regulatory measures, and community engagement.
HIGHLIGHTS
A real-time multi-sensor probe was used to measure the physical–chemical parameters in marine water.
Temperature, conductivity, pH, DO, turbidity, and depth were measured directly.
Total dissolved solids and salinity were calculated through built-in correlations.
Physical–chemical parameters provided valuable information on water quality.
The measured parameters revealed good water quality in the Durrës Bay.
INTRODUCTION
Coastal areas are the most developed regions, characterized by a high population and a wide range of industrial and agricultural activity, which puts them under high pressure and pollution from various natural and anthropogenic hazards. Those have large social, economic, and environmental values and are pointed out as the most productive zones that provide social and economic benefits to humans (Melet et al. 2020). Among natural hazards, the extreme heat waves that cause fish mortality, large sea waves and changes in sea levels, erosion and salinization of aquifers, low oxygen content, and acidic degradation are dangerous phenomena (Melet et al. 2020). At the same time, anthropogenic activities like maritime pollution, eutrophication, overfishing, degradation or loss of marine and coastal ecosystems and habitats (Melet et al. 2020), and uncontrolled urban and industrial discharges degrade the water quality. The quality of coastal water plays an important role in the sustainable development of coastal areas, particularly from the point of view of aquatic life and tourist activity. Coastal areas are very important as they have been associated with depressive tendencies in sea surface warming trends over the past few decades. This association is related to the ability of coastal areas to slow down global warming trends and reduce the extreme events of marine heatwaves (Marin et al. 2021). For this reason, monitoring and understanding marine coastal environmental status and hazards are of increasing interest. A variety of instruments and sensors have been developed and have been used to monitor the characteristic parameters of coastal water essential for indicating the pollution level.
Seawater monitoring represents a fundamental component in the field of marine science. It offers a systematic and data-based approach for comprehending the complex interactions within aquatic ecosystems. The significance of monitoring seawater extends far beyond the scientific realm, as it underpins our ability to comprehensively assess the health of marine ecosystems, track climate change effects, and manage vital coastal resources. Monitoring and assessing marine water quality is a paramount concern due to its implications for public health, ecosystem sustainability, and the economic well-being of the region.
The Mediterranean Sea is the most densely populated closed sea in the world. This concentration is intensifying year after year and generating more pollution and disturbance, leading to environmental degradation and increased risks for coastal populations and infrastructure (UNEP/MAP & Plan Bleu 2020). The Mediterranean Sea has approximately 150 million inhabitants on its coasts. In addition to numerous tourists, high levels of industrial and shipping activities are causing a rapid increase in marine pollution. This is combined with other anthropogenic drivers of environmental change, including climate change (e.g., seawater temperature, heatwaves, salinity, acidification, extreme events, and sea-level rise), unsustainable land- and sea-use practices, and non-indigenous species (Ziveri et al. 2023). Intensive human activities along the coastlines of the Mediterranean Sea increase the pressure on its marine environment, making it one of the most polluted sea environments in the world (Karadirek et al. 2019). Pollution in this area can harm aquatic life, degrade water resources, and undermine the very foundations of a thriving maritime economy. Therefore, understanding the main pollution sources, trends, and impacts is essential for formulating effective mitigation strategies and safeguarding the Mediterranean environment.
Marine pollution from the growing urbanization of coastal areas has increased the number of land- and sea-based pollution sources, including shipping and the exploitation of marine resources. The effects of growing marine and coastal tourism have caused different negative impacts, like increasing amounts of untreated sewage and waste, the degradation and loss of coastal habitats, and a loss of biodiversity (World Bank Report 2020).
Rapid urban developments have affected the cleanliness of the Albanian coast. The Adriatic and Ionian Seas have become hosts for urban, industrial, agricultural, and livestock discharges. Numerous spills occur from pesticides, raw chemicals from agricultural lands, organic waste containing phosphorus and nitrogen, viruses and pathogenic bacteria, heavy metals, etc. In addition to this, the increased number of inhabitants in the urban centers has made the process of self-cleaning the sea impossible (WFD and Eco Albania 2021).
The importance of marine water monitoring in mitigating marine incidents lies in its ability to quickly detect and prevent problems such as oil spills, waste dumping, thermal pollution, and harmful algal blooms. This facilitates a prompt reaction to such incidents, reducing their negative effects on marine ecosystems, coastal communities, and the economy. Although there has not been any major maritime oil spill incident within the Mediterranean region, accidents are considered inevitable occurrences, and the risk of one happening in the near future cannot be ruled out (Bellefontaine et al. 2016). The research data presented in this paper were obtained during the implementation of the Interreg ADRION project SEAVIEWS (SEctor Adaptive VIrtual Early Warning System for Marine Pollution), aiming at the development of a transnational repository network that receives, stores, and analyzes data about seawater quality from a network of smart sensors installed in the Adriatic and Ionian Seas (Interreg ADRION n.d.). The core of this network is the innovative web-based platform that reads, stores, and analyzes the input data from three main categories: data from the smart sensor network allocated in critical points in the ADRION area; data from the users of developed mobile application (photographs, videos, and GPS coordinates related to oil spills, marine litter, industrial discharges, or other relevant observations); and data from available databases in the area. The analyses applied to the data within the platform include descriptive statistics, which involve calculating statistical measures as a basis for more advanced types of analyses. In addition, big data analytic techniques are implemented, utilizing smart algorithms for managing and discovering patterns and relationships within large datasets, which may otherwise be overlooked or ignored. By leveraging these analyses, the web-based platform aims to uncover hidden patterns, provides a solid foundation for data-driven decision-making, and helps facilitate proactive measures to address environmental pollution in the ADRION area. The SEAVIEWS web platform enables different levels of access and involvement for the marine stakeholders, like Marinas and Port authorities, other marine organizations, civil emergencies authorities, NGOs, and so on.
In this paper, we will utilize comprehensive monitoring data, scientific analysis, and the latest research findings to evaluate the state of water quality in the area of the Port of Durrës, on the Albanian Adriatic coast, offering valuable insights for decision-makers, environmental authorities, and port operators.
MATERIALS AND METHODS
Monitoring site
Durrës Bay is about 18 km long from north to south, with a coastline of about 20 km to the east. To the west, the waterline is more than 10 m deep. The Durrës Bay is well protected by the Durrës Cape, which provides shelter from the east through to the northwest, but the main breakwater, which was built in a southeastern direction from the shore, extends that shelter through to the south. The City of Durrës lies in the geographical position latitude 41°19′ North and longitude 19° 27′ East. It is about 35 km away from the capital Tirana, 300 km from the port of Bari, and 200 km from Brindisi, Italy. Durrës city has 292,029 inhabitants (INSTAT 2021) and represents the most important transport, maritime, road, and touristic hub in the country (WBG 2016). Most environmental problems on the coast are caused by poor wastewater treatment, the lack of effective municipal waste collection and recycling, and intense port activities.
The Port of Durrës is an important maritime hub that serves not only as a crucial gateway for the Albanian economy but also plays a pivotal role in the broader context of Adriatic maritime commerce. It is the principal port in Albania, handling roughly 90% of the country's international maritime trade tonnage and 85% of all the country's export and import trade. As a vital interface for the country on the Adriatic coast, linking Albania with other Mediterranean and Balkan countries, it opens up the country to an additional market of 40 million people. The current level of traffic is about 3.8 million tons of cargo per year and approximately 80,000 passengers (Metalla et al. 2016). The ferry terminal in Durrës is the gateway to Albania and the Balkan region, with ferry connections to Italy. More than 850,000 passengers, 185,000 cars, and 76,000 cargo units pass the terminal annually (AFTO n.d.). Seaports and terminals are major hubs of economic activity and major sources of pollution. Port operations can cause significant damage to water quality and, subsequently, to marine life and ecosystems. These effects may include contamination of commercial fish and shellfish, depletion of oxygen in water, and bioaccumulation of certain toxins in fish. Major water quality concerns at ports include wastewater, the leaking of toxic substances from ships, storm water runoff, and dredging. The presence of polycyclic aromatic hydrocarbons and other priority organic pollutants in the waters of Durrës Bay has been reported recently (Halo et al. 2023).
Nowadays, the Durrës Bay faces unique challenges in urban development and water pollution. The surge in tourism and population growth has led to increased pressure on the bay's urban infrastructure. Inadequate waste disposal systems and outdated sewage treatment facilities contribute to elevated levels of pollutants entering the water. The bay's delicate ecosystem faces threats from industrial and port activities, as well as from urban expansion. Addressing these issues requires a holistic approach, integrating modernized infrastructure, stringent environmental regulations, and community engagement to ensure the sustainable development of the area. Recently, the construction of the Yacht Marina was started in Durrës. This project is expected to bring economic growth to the area and a further development of tourism.
Measuring equipment
The HYDROLAB HL7 multiparameter sonde was chosen for seawater measurements. This sophisticated instrument can thrive in demanding environmental conditions for long-term continuous monitoring. It maximizes deployment lifespan, lowers maintenance needs, and provides traceable data for high-quality, reliable monitoring of vital changes in water quality. The Operating Software streamlines data collection and calibration tasks necessary to validate accurate data. The battery life is 90 days, the maximum measuring length is 200 m, and the measuring frequency can arrive at one measurement per second. The sonde helps environmental scientists to correctly log data autonomously and integrate it into real-time telemetry systems. The equipment can be adjusted for static measurements as well as for use on moving vessels, offering increased flexibility in monitoring campaigns.
The HL7 sonde, with a large sensor suite, is able to thrive in demanding environmental conditions for long-term continuous monitoring. Bio-fouling is minimized when equipped with the central cleaning brush, and performance is maximized with an advanced power management system (OTT HydroMet n.d.). The HL7 sonde held the sensors for temperature, conductivity, pH, dissolved oxygen (DO), turbidity, and depth. Other parameters, such as salinity, total dissolved solids (TDS), and oxidation–reduction potential, were calculated through built-in correlations. The data logger connected to the sonde was programmed to acquire data and send them every 15 min to the server.
Data curation and analysis
This research is part of the SEAVIEWS Project, which aims to enhance the ability to address environmental vulnerability, fragmentation, and the protection of ecosystem services in the Adriatic Sea. This is enabled by using a network of smart sensors allocated at critical points and supporting the establishment of digital labs to be used as channels for information, data analysis, and research results to be circulated to the general public. In our study, data analysis was carried out through statistical analysis.
Descriptive statistical analysis was used to check the concentration level and variability of the investigated parameters. Pearson's Rho correlation was applied to investigate the associations between parameters (p < 0.05) and then factor analysis (FA) was used to clarify the association between parameters and discuss the probable sources of their association. FA groups the investigated parameters into different factors based on their similarity level (Pearson correlations). FA is an important tool for extracting useful information from the investigated data after grouping the variables into separated factors. Time series are used to visualize the distribution of the data during the full monitoring period and to compare the measured data with the recommended values of each parameter.
The measured data for each parameter were transmitted every 15 min to the computer for the entire period under study, December 2021–November 2022. It was observed that in about 10.8% of the total number of measurements (N = 3,772 out of 35,040 measurements), the sonde did not transmit data for any of the parameters in the study. For some of the parameters, such as TDS, salinity, and turbidity, this number was higher. For this, the correlation between the parameters was studied, and it turned out that between electro-conductivity and TDS, as well as salinity, there was a very strong correlation (r ≈ 1, p = 0.000). Based on it, the linear regression between TDS–electrical conductivity (EC) and salinity–EC was studied.
TDS is an important parameter for the characterization of natural waters for environmental, geochemical, and petrochemical studies. The method of determining TDS through linear correlation with the EC of groundwater, surface water, or marine water systems has found wide use (Rebello et al. 2020) due to its advantages over the gravimetric method. It consists in the weighing of the dry residue obtained from evaporating a certain volume of a filtered water sample (APHA/AWWA/WEF 1995).
Regression analysis . | Regression equation . | Model summary . | ||
---|---|---|---|---|
TDS (g/L) versus EC (mS/cm) | TDS = −0.00391 + 0.6401 EC | S | R2 | R2 (adj) |
0.0289 | 99.99% | 99.99% | ||
Salinity (PSU) versus EC (mS/cm) | Salinity = −2.892 + 0.7133 EC | S | R2 | R2 (adj) |
0.0980 | 99.92% | 99.92% |
Regression analysis . | Regression equation . | Model summary . | ||
---|---|---|---|---|
TDS (g/L) versus EC (mS/cm) | TDS = −0.00391 + 0.6401 EC | S | R2 | R2 (adj) |
0.0289 | 99.99% | 99.99% | ||
Salinity (PSU) versus EC (mS/cm) | Salinity = −2.892 + 0.7133 EC | S | R2 | R2 (adj) |
0.0980 | 99.92% | 99.92% |
Variable . | N . | N* . | Mean ± SD . | CV% . | Min . | Q1 . | Median . | Q3 . | Max . | Skewness . | Kurtosis . |
---|---|---|---|---|---|---|---|---|---|---|---|
DO (mg/l) | 31,252 | 16 | 8.1 ± 0.796 | 10 | 5.7 | 7.6 | 8.1 | 8.6 | 12 | 0.01 | −0.11 |
TDS (g/l) | 31,263 | 5 | 27 ± 3.2 | 12 | 10.3 | 25 | 27 | 29 | 33 | −0.64 | 0.96 |
EC (mS/cm) | 31,263 | 5 | 42 ± 4.9 | 12 | 16.2 | 39 | 42 | 46 | 52 | −0.64 | 0.96 |
Salinity (PSU) | 31,263 | 5 | 27 ± 3.5 | 13 | 9.5 | 25 | 27 | 30 | 34 | −0.55 | 0.65 |
pH | 31,262 | 6 | 8.8 ± 0.08 | 1 | 8.5 | 8.7 | 8.8 | 8.8 | 9.1 | 0.09 | −0.17 |
Turbidity (NTU) | 25,805 | 5,463 | 221 ± 400 | 181 | 0 | 26 | 52 | 179 | 2,000 | 2.63 | 6.54 |
Water Temp (°C) | 31,266 | 2 | 18.7 ± 4.7 | 25 | 10.3 | 14.1 | 19 | 23 | 29 | 0.11 | −1.22 |
Depth (m) | 31,252 | 16 | 3.0 ± 0.125 | 4 | 2.63 | 2.95 | 3.0 | 3.1 | 3.6 | 0.28 | 055 |
Variable . | N . | N* . | Mean ± SD . | CV% . | Min . | Q1 . | Median . | Q3 . | Max . | Skewness . | Kurtosis . |
---|---|---|---|---|---|---|---|---|---|---|---|
DO (mg/l) | 31,252 | 16 | 8.1 ± 0.796 | 10 | 5.7 | 7.6 | 8.1 | 8.6 | 12 | 0.01 | −0.11 |
TDS (g/l) | 31,263 | 5 | 27 ± 3.2 | 12 | 10.3 | 25 | 27 | 29 | 33 | −0.64 | 0.96 |
EC (mS/cm) | 31,263 | 5 | 42 ± 4.9 | 12 | 16.2 | 39 | 42 | 46 | 52 | −0.64 | 0.96 |
Salinity (PSU) | 31,263 | 5 | 27 ± 3.5 | 13 | 9.5 | 25 | 27 | 30 | 34 | −0.55 | 0.65 |
pH | 31,262 | 6 | 8.8 ± 0.08 | 1 | 8.5 | 8.7 | 8.8 | 8.8 | 9.1 | 0.09 | −0.17 |
Turbidity (NTU) | 25,805 | 5,463 | 221 ± 400 | 181 | 0 | 26 | 52 | 179 | 2,000 | 2.63 | 6.54 |
Water Temp (°C) | 31,266 | 2 | 18.7 ± 4.7 | 25 | 10.3 | 14.1 | 19 | 23 | 29 | 0.11 | −1.22 |
Depth (m) | 31,252 | 16 | 3.0 ± 0.125 | 4 | 2.63 | 2.95 | 3.0 | 3.1 | 3.6 | 0.28 | 055 |
RESULTS AND DISCUSSION
Table 2 shows that the sensor for turbidity measurement did not function and did not transmit data to the server for a period of 48 days (N* = 5,463 measurements). Other sensors functioned normally (N* = 2–16 measurements). All parameters in the study, with the exception of turbidity, were quite stable for the entire period under study, with a very low variation (CV%: 1–25% ≤ 25%) and a narrow range of data fluctuation (skewness and kurtosis are close to 0).
The boxplot diagrams revealed a relatively stable situation for pH, depth, and water temperature parameters. It is supported by descriptive statistical parameters, which revealed narrow ranges and symmetric outliers of their respective data on both sides of minimum and maximum values (8.8 ± 0.08 for pH, 3.0 ± 0.125 for depth, and 18.7 ± 4.7 for water temperature). Higher outlier values greater than the median concentration were detected for DO, and EC, TDS, and salinity revealed the data are skewed, leaving a high number of outlier values lower than their respective median values. Turbidity is the sole parameter that shows very high variation (CV% = 181% > 75%). It revealed a high concentration level and a great number of outlier data, higher than the median value, indicating that the data are skewed right.
The time series of the measured data show a temporal trend of the investigated parameters measured for 12 months, from December 2022 to November 2023. A few spikes in data caused by spontaneous factors (6.6% of the total data for the depth values and less than 0.27% of the total data for other parameters) do not affect the values of descriptive statistical data, obtained from 35,040 measurements for all parameters.
Water temperature
Temperature affects water parameters such as solubility and chemical equilibria of oxygen, gases, and other chemicals in water. It has a strong effect on the biodegradation process of organic material in water and sediment, which increases the oxygen demand and affects the DO level. The measured temperature ranged from 10.3 to 29 °C. It revealed a symmetrical distribution of Q1 (14.1 °C) and Q3 (23 °C) around the median temperature value (19 °C) (Table 2). Water temperature showed seasonal variation, with lower values in the winter and higher values in the summer (Figure S1 in the Supplementary Material). The water temperature of the Adriatic Sea during 2019 ranged from 13 °C in February and up to 27 °C in August.
In the northern Adriatic Sea, the surface temperature near the coast ranges from about 5 °C in winter to 27 °C in summer, with a difference greater than 20 °C observed between winter and summer (Russo & Artegiani 1996). The temperature values recorded in the Adriatic Sea at the Durrës site (10.3–29 °C) match with the temperatures recorded in the southern Adriatic Sea, higher than 13.5 °C (Russo & Artegiani 1996). Higher temperatures (10.3–29 °C) were recorded during 2023 at the Durrës site compared with those recorded in the northern Adriatic Sea (5–27 °C). It is affected by the geographical position of Durrës site and the trend of temperature increases due to global climatic changes, with a mean value of 1.27 °C during the 35-year study of the period 1982–2016 (Pastor et al. 2018) and about 1 °C in southern Albania since 1970 (Knez et al. 2022). Compared with the surface water temperature of Durrës coast published by Gjiknuri (1995), the minimum value increased by 2.6 °C, from 7.7 °C in 1995 to 10.3 °C in 2022, while the maximum value remained the same.
TDS, salinity, and EC
TDS, salinity, and EC data revealed the same temporal trend over the full monitoring period, indicating a strong relationship between them. The measured TDS, salinity, and EC ranged within a narrow interval (10.3–33 g/L for TDS, 9.5–34 g/L for salinity, and 16.2–52 mS/cm for EC) and revealed a symmetrical distribution around median values (Table 2).
Salinity is an important parameter of seawater quality because it has a strong effect on the marine biota. The South Adriatic Sea shows high salinity (>38.6 PSU) and low variability (Lipizer et al. 2014). A similar situation was observed at the Durrës site, with Q1–Q3 recorded from 25 to 30 PSU and a mean value of 27 ± 3.5 PSU. Some fluctuations of lower salinity were observed during the rainy seasons of spring and autumn. A higher salinity change was recorded in the Eastern Mediterranean Sea (Emeis et al. 2000).
EC data show a constant situation during the full monitoring period, evaluated by low variability (CV% = 12% < 25%) and low kurtosis (K = 0.96 < 3). This indicates a narrow range of EC (mean ± SD =42 ± 4.9) and a relatively stable water composition. Siosemarde et al. (2010) mentioned the presence of many cases of reasonably constant water composition in a given region or study site.
pH
The pH of aquatic ecosystems depends on the chemical and biological activity of the water. In general, pH is relatively stable. The changes in pH values are usually caused by anthropogenic pollution, photosynthesis, or the respiration of algae and bacteria. Most ecosystems are sensitive to changes in pH, and the monitoring of pH has been incorporated into the marine water standards (PHILMINAQ 2008). On the other hand, under the great effect of the carbonate–bicarbonate buffer system in seawater, the pH of the water remains quite stable (Trang et al. 2020).
Dissolved oxygen
Low DO values (DO > 5.7 mg/L) could be explained by the effects of various natural and anthropogenic factors. In coastal areas, where the physical processes are generally dynamic and complex, DO concentration depends on multiple factors. Some of them are the hydrological conditions affecting gas solubility, air–water exchange, water vertical stratification, and pelagic and benthic metabolism, where the net balance between oxygen production and consumption processes is a key factor affecting changes in DO concentration in coastal waters (Kralj et al. 2019). Water ecosystems can become undersaturated with oxygen when natural processes and/or anthropogenic processes produce enough organic carbon that is aerobically decomposed faster than the rate of oxygen deaeration (Rabalais et al. 2010). The Durrës Bay is affected by two Albanian rivers, Ishmi and Erzeni, the most polluted rivers in Albania, which discharge different pollutants in the bay. The positive temperature trend in bottom waters, coupled with the increase in riverine discharges in late spring, limiting vertical mixing and bottom water renewal, may favor events of oxygen depletion in coastal ecosystems (Kralj et al. 2019).
Water upwelling is another process affecting DO due to the heterotrophic processes in coastal environments. These processes lower the DO during the degradation of organic matter in the water column or bottom sediments, using DO during the oxidation process of reduced constituents such as sulfide and methane in water (Zhang et al. 2010). Besides this, coastal upwelling can bring high concentrations of nutrients and sinking organic particles to surface waters. It can stimulate oxygen production and also bring low DO due to local or large-scale oxygen demand from the microbial decay of sinking organic particles (Zhang et al. 2010). These values are within an acceptable range for aquatic life, as the optimal DO values are higher than 5 mg/L (Zhang et al. 2020). Similar DO values (mean value of 8.7 mg/L) were registered at the Vlora Bay in May 2014 measured at 0.5 m deep (Kane et al. 2015) at 14 sampling sites positioned 100 m far from the coast. The mean DO value of 8.6 mg/L was also registered in the Durrës Bay (Figure S1 and Table S1 in the Supplementary Material).
Seawater depth
Turbidity
Time series analysis
The data of this study were measured from a single site, transmitted to the server at 15-min time intervals, and found to be relatively stable (CV% ≤ 25%) for all parameters except turbidity. Under such conditions, time series were used to visualize the distribution of the data during the full monitoring period and to compare the measured data with the recommended values of each parameter. The time domain analysis is also performed by investigating the stability of the dataset for each parameter studied by the linear trend model.
The mean absolute percent error (MAPE), which expresses the accuracy as a percentage of the error, is used to compare the fits of different time series models (in our case the linear model) with the measured data. The measured data close to zero greatly inflates to very high MAPE values, because it is calculated as the ratio of the absolute error to the actual data. In this case, it is better to use the MAPE data in combination with the respective CV% values (Table 3).
Parameters . | CV% . | MAPE (%) . | Linear equation models . |
---|---|---|---|
DO | 10 | 6.6 | DO = 8.72 + 0.00004t |
EC | 12 | 9.2 | EC = 44.7 − 0.000165t |
pH | 12 | 0.7 | pH = 8.76 + 0.000002t |
Salinity | 13 | 10 | Salinity = 29.1 − 0.000124t |
Seawater depth | 4 | 3.4 | Depth = 3.05 + 0.000001t |
TDS | 12 | 9.2 | TDS = 26.6 − 0.000106t |
Turbidity | 181 | 754 | Linear model not available |
Water temperature | 25 | 14 | Water Temp = 12.5 + 0.00039t |
Parameters . | CV% . | MAPE (%) . | Linear equation models . |
---|---|---|---|
DO | 10 | 6.6 | DO = 8.72 + 0.00004t |
EC | 12 | 9.2 | EC = 44.7 − 0.000165t |
pH | 12 | 0.7 | pH = 8.76 + 0.000002t |
Salinity | 13 | 10 | Salinity = 29.1 − 0.000124t |
Seawater depth | 4 | 3.4 | Depth = 3.05 + 0.000001t |
TDS | 12 | 9.2 | TDS = 26.6 − 0.000106t |
Turbidity | 181 | 754 | Linear model not available |
Water temperature | 25 | 14 | Water Temp = 12.5 + 0.00039t |
The MAPE (%) data of all parameters, except turbidity, are smaller than 14% and followed by small variation (CV% ≤ 25%). It means the data of DO, pH, EC, TDS, salinity, depth, and water temperature are stable. It is confirmed by small slopes of the linear equations model (b = 0.000001–0.00004 for the depth, pH, and DO; b = 0.00011–0.00017 for TDS, salinity, and EC; and b = 0.00039 for water temperature). Higher slopes, CV%, and MAPE values indicate higher effects of seasonal conditions or other factors.
Multivariate analysis
The measured physical–chemical data were analyzed by Pearson's correlation analysis (Table 4). The correlation matrix data indicate the level of similarity between particular physical–chemical parameters. At least five similar pairs of parameters were found, characterized by a correlation coefficient r > 0.4 and p = 0. Very strong positive correlations were found between the pairs of EC–TDS, EC–salinity, and TDS–salinity data (Table 4), followed by a moderate and significant negative correlation (r > 0.4, p = 0.000) between DO and water temperature, and a weak and significant correlation between pH and DO (r = 0.317, p = 0.000) (Table 4).
Variables . | DO . | TDS . | EC . | Salinity . | pH . | Turbidity . | Water Temp . |
---|---|---|---|---|---|---|---|
TDS | 0.068 | ||||||
EC | 0.068 | 1.000* | |||||
Salinity | 0.078 | 1.000* | 1.000* | ||||
pH | 0.317* | 0.216 | 0.216 | 0.211 | |||
Turbidity | −0.206 | −0.126 | −0.126 | −0.131 | 0.041 | ||
Water Temp | − 0.418* | −0.206 | −0.206 | −0.228 | 0.268 | 0.243 | |
Depth | −0.152 | −0.027 | −0.027 | −0.027 | −0.280 | 0.034 | −0.010 |
Variables . | DO . | TDS . | EC . | Salinity . | pH . | Turbidity . | Water Temp . |
---|---|---|---|---|---|---|---|
TDS | 0.068 | ||||||
EC | 0.068 | 1.000* | |||||
Salinity | 0.078 | 1.000* | 1.000* | ||||
pH | 0.317* | 0.216 | 0.216 | 0.211 | |||
Turbidity | −0.206 | −0.126 | −0.126 | −0.131 | 0.041 | ||
Water Temp | − 0.418* | −0.206 | −0.206 | −0.228 | 0.268 | 0.243 | |
Depth | −0.152 | −0.027 | −0.027 | −0.027 | −0.280 | 0.034 | −0.010 |
*p = 0.000.
Natural waters are complex systems characterized by various ionic ratios of specific ions and the average activity of all ions present in water, which cause a nonlinear relationship between TDS and EC in water systems (Hubert & Wolkersdorfer 2015). The investigated parameters were continuously measured at the same sampling site, which exhibits similar conditions in the measured parameters, and much higher contents of Na and Cl ions than other characteristic ions of the seawater. It establishes stable ionic strength and ionic ratios, which are favorable parameters for establishing a linear relationship between EC and TDS. Under such conditions, very strong and significant correlations (r = 1, p = 0.000) were found between the pairs EC–TDS and EC–salinity.
FA is performed to better explain the associations between the measured parameters. The results of FA are shown in Table 5.
Variable . | Factor 1 . | Factor 2 . | Factor 3 . | Communality . |
---|---|---|---|---|
EC (mS/cm) | 0.993 | 0.000 | 0.000 | 0.998 |
TDS (g/l) | 0.993 | 0.000 | 0.000 | 0.998 |
Salinity (PSU) | 0.992 | 0.000 | 0.000 | 0.998 |
Water Temp (°C) | 0.000 | −0.827 | 0.000 | 0.767 |
DO (mg/l) | 0.000 | 0.698 | −0.479 | 0.718 |
Turbidity (NTU) | 0.000 | −0.597 | 0.000 | 0.365 |
pH | 0.000 | 0.000 | −0.841 | 0.792 |
Depth (m) | 0.000 | 0.000 | 0.682 | 0.474 |
Variance | 3.026 | 1.600 | 1.483 | 6.109 |
% Var | 0.378 | 0.200 | 0.185 | 0.764 |
Variable . | Factor 1 . | Factor 2 . | Factor 3 . | Communality . |
---|---|---|---|---|
EC (mS/cm) | 0.993 | 0.000 | 0.000 | 0.998 |
TDS (g/l) | 0.993 | 0.000 | 0.000 | 0.998 |
Salinity (PSU) | 0.992 | 0.000 | 0.000 | 0.998 |
Water Temp (°C) | 0.000 | −0.827 | 0.000 | 0.767 |
DO (mg/l) | 0.000 | 0.698 | −0.479 | 0.718 |
Turbidity (NTU) | 0.000 | −0.597 | 0.000 | 0.365 |
pH | 0.000 | 0.000 | −0.841 | 0.792 |
Depth (m) | 0.000 | 0.000 | 0.682 | 0.474 |
Variance | 3.026 | 1.600 | 1.483 | 6.109 |
% Var | 0.378 | 0.200 | 0.185 | 0.764 |
Three main factors were extracted from factor loads (Table 4). Factor 1 is compiled by very high loads of EC, TDS, and salinity, all derived from the same factor in seawater – the content of the dissolved ions. Factor 2 is compiled by a high negative load of water temperature and a positive load of DO, which indicate a reverse relationship between temperature and DO. This is a normal phenomenon suggesting that temperature changes contribute to the surface DO distribution, such as increasing solubility in cold seasons and degassing in warm seasons, which likely represents natural components or climate processes (Lee et al. 2023). The association of the turbidity in this factor with a moderate negative load is not clear. Factor 3 is compiled by a high negative load of pH and a positive load of depth, as well as a moderate negative load of DO (Table 4). The pH and DO change in the same sense, indicating the DO content is higher in clean water than in acidic conditions. A similar phenomenon was investigated by Lee et al. (2023). Another finding from Factor 3 is related to the inverse relationship of pH and DO with depth, indicating that pH and DO are lower at higher depths and higher in surface.
CONCLUSIONS
In this study, the seawater quality from the Durrës site, Albania, was assessed using an automatic measuring sonde for the measurement of the physical and chemical parameters. The obtained data for all parameters, except turbidity, were within the recommended values for the survival of aquatic life.
Low variability (CV% < 25%) on the measured parameters, except turbidity, indicates a stable status in the water quality from the Durrës site. Time series revealed small seasonal variations on all parameters, except turbidity.
The measured data confirmed the linear relationship between EC and TDS, which enables the calculation of TDS based on the conversion factor from the equation of their linear relationship. Since the measurements were carried out under a low variation of the parameters measured at the same monitoring site, it has favored a fairly high linear regression between EC and TDS (r = 1), which provided a correct calculation of TDS values.
TDS, salinity, and EC data revealed the same temporal trend over the full monitoring period, indicating a strong relationship between them. Salinity is an important parameter of seawater quality, because it has a strong effect on the marine biota. The Durrës site revealed a stable situation characterized by a very close Q1–Q3 range, which was recorded from 25 to 30 PSU with a mean value of 27 ± 3.5 PSU. Some fluctuations of lower salinity were observed during the rainy seasons of spring and autumn.
The high negative load of water temperature and the positive load of DO in Factor 1 revealed a reverse relationship between them. This indicates that the solubility of DO decreased with the increase in temperature, which is a normal phenomenon in water. The association of the turbidity in this factor with a moderate negative load is not clear. The pH and DO change in the same sense, indicating that the DO content is higher in clean water than in acidic water conditions. pH and DO show inverse relationships with depth, indicating that pH and DO are lower at higher depths.
The small variations of full data and monthly data, smaller than those of early data, indicate the effectiveness of monthly or seasonal monitoring programs. Nonetheless, the sporadic peaks observed at outlier points indicate the need for conducting repeated in situ measurements until stable outcomes are achieved. While the obtained data for physical–chemical parameters in this study align with the recommended values, it is important to prioritize the maintenance and enhancement of water quality in the Durrës Bay. This can be done through regular monitoring of marine waters and by including other chemical parameters in future studies.
This study is very important for understanding the current state of marine water quality in the Durrës Bay. It addresses the gaps in existing data on the physical–chemical parameters in the area and provides an essential baseline for tracking changes over time and assessing the impact of human activities on the marine environment. Sharing the results of this study internationally can foster collaboration and knowledge exchange, for a more comprehensive understanding of the Adriatic Sea's water quality and implementation of regional environmental management strategies.
Decision-making bodies, including environmental agencies and policymakers, can leverage these research findings to enhance systematic monitoring programs for real-time data and prompt responses to emerging issues. These entities can develop evidence-based policies, enforce strict regulations on industrial and municipal discharges to mitigate pollution, advocate for sustainable coastal development, implement erosion control measures, establish comprehensive waste management programs in coastal communities to prevent marine pollution, and conduct educational campaigns to increase awareness of individual impacts on marine water quality. These actions are essential to mitigate human-induced stress on the ecosystem, safeguard critical habitats, and contribute to the overall preservation of the Durrës Bay's aquatic environment.
FUNDING
This research was funded by the European transnational programme INTERREG V-B Adriatic-Ionian Programme ADRION 2014-2020, under the project Sector Adaptive Virtual Early Warning System for marine pollution (SEAVIEWS) (Project No. ADRION-951).
AUTHOR CONTRIBUTIONS
EM contributed to the conceptualization, methodology, formal analysis, and writing and revising the original draft. JT contributed to formal analysis and investigation. PL contributed to the conceptualization, data analysis, and writing and revising the original draft. SD contributed to the methodology and formal analysis. FQ, AN, and BM contributed to formal analysis. All authors have read the draft of the manuscript and approved it.
DATA AVAILABILITY STATEMENT
Data cannot be made publicly available; readers should contact the corresponding author for details.
CONFLICT OF INTEREST
The authors declare there is no conflict.