The study was carried out in Vlora Bay, Albania, a coastal area historically affected by mercury releases from a former chlor-alkali and PVC industrial complex. Seven stations were selected to represent locations with different expected contamination levels. Stations S1-S5 were located within Vlora Bay, whereas S6 (Porto Palermo) and S7 (Butrint Lagoon, Saranda) were used as reference stations. Figure 1 shows the location of the sampling stations, defined as follows: S1, Petrolifera, directly in front of the former chlor-alkali plant; S2, Military Marine Academy, where only Patella specimens were collected; S3, Kristera Hotel, near the main drinking-water intake area of Vlora; S4, Bova, a buoy station close to S3; S5, Karaburun Peninsula, a comparatively undisturbed military-restricted area; S6, Porto Palermo; and S7, Butrint Lagoon, where mussels were farmed under natural conditions.

Figure 1.

Sampling stations

The sampling campaign was conducted during 2023 and 2024. Specimens of Patella caerulea and Mytilus galloprovincialis were collected manually from rocky coastal sites and placed in plastic bags containing seawater for transport to the laboratory. The samples were collected randomly at each station to provide representative coverage of the local population. More than 50 individuals of Patella caerulea were analysed from six stations (S1-S6). At station 7, Patella caerulea has not been found. Limpets from the genus Patella typically inhabit rocky coastal environments. In the Mediterranean Sea, this genus includes four species: P. caerulea, P. ulyssiponensis, P. rustica, and P. ferruginea [32]. P. caerulea (Linnaeus, 1758), endemic to the Mediterranean, is one of the most common species found on rocky shores [33]. In addition, intraspecific genetic variation has been reported in Mediterranean P. caerulea in response to changing hydrodynamic conditions [34]. Its presence extends to the rocky areas of Vlora Bay and the Ionian Sea coast.

A total of 33 individuals of Mytilus galloprovincialis were collected from six stations (S1 and S3-S7); mussels were not found at station S2. The collected Mytilus galloprovincialis specimens ranged from 2.9 to 4.7 cm in shell length, whereas Patella caerulea ranged from 1.3 to 4.3 cm. In the laboratory, samples were frozen at -20 C until analysis. Only the soft tissue of each individual was used for mercury determination; no pooled or composite samples were prepared.

Fish samples were obtained from local fishermen in Vlora, Durres, and Saranda during 2023-2024. Fifty four individuals representing 12 species were analysed. Each fish was identified before dissection, and muscle tissue was separated from the bone while keeping the skin attached to the muscle. Tissue samples were placed in polyethylene bags and immediately frozen. All biological samples were analysed for total mercury on a wet-weight basis.

In addition, two series of soil-core samples were collected from the remediated industrial area: one from the former electrolysis plant and one from the former catalytic plant. These samples were used to evaluate residual mercury in the terrestrial component of the former industrial hotspot.

All reagents used in this study were of high-purity grade (Suprapur, Merck, Germany), and Milli-Q water was used throughout the analytical procedure. Mercury concentrations were determined with a laboratory-constructed cold-vapour atomic absorption system coupled to an Analytik Jena Model 800 F instrument (Germany).

Validated digestion procedures were applied for both biota and soil in accordance with the Marine Environment Studies Laboratory standard operating procedures [35]. Approximately 1.0-1.5 g of each biological sample was digested with concentrated nitric acid. To ensure complete oxidation of mercury species to the inorganic form, the oxidation step was strengthened by adding bromine monochloride or dichromate ions. Mercury was quantified by absorbance spectroscopy using the standard-addition method.

During digestion, nitrogen oxides produced spectral interference even when deuterium-background correction was used. This interference was minimised by bubbling air through the reducing cell for 1 min before the addition of the reducing agent. Soil samples were digested and analysed by the same analytical principle [35].

Analytical accuracy was checked using the certified reference material IAEA-407 Fish. Recovery values for mercury ranged from 83% to 96%, confirming acceptable analytical performance. Procedural blanks showed no evidence of contamination. The use of individual samples, certified reference materials, blanks, and matrix compensation via the standard-addition method enhanced the reliability and repeatability of the measurements.

Descriptive statistics were calculated for each species and station. Before analysis of variance, the assumptions of normality and homogeneity of variances were assessed using the Shapiro-Wilk and Levene tests, respectively. Differences among stations or species were evaluated by one-way analysis of variance, followed by Tukey's honest significant difference post hoc test when the overall test was significant. Statistical significance was defined at p < 0.05. Statistical analyses were performed using StatSoft Statistica 12.

Two core soil samples were collected from the areas historically identified as the most contaminated zones of the former electrolysis plant (series 1, Figure 2) and the catalytic plant (series 2, Figure 2), both located within the former chlor-alkali industrial complex that has undergone remediation in recent years. These locations were selected because they were previously associated with intensive industrial activities and were therefore considered potential hotspots of mercury contamination. The measured mercury concentrations in the analysed samples ranged from 0.023 to 0.076 µg/g.

Figure 2.

Mercury (Hg) concentration in the soil of the former Soda Chlor-Alkali Plant.

These values are relatively low and indicate that the remediation measures implemented at the site have been highly effective in reducing residual mercury contamination in surface soils [6]. Numerical sediment quality guidelines (SQGs) for freshwater ecosystems have previously been developed [36]. In our study, to better evaluate the environmental significance of these concentrations, the results were compared with international sediment quality guidelines. In particular, the measured values were assessed against the Threshold Effects Level (TEL = 0.13 µg/g) established by the National Oceanic and Atmospheric Administration (NOAA). All measured concentrations were well below the TEL value, suggesting that the current levels of mercury in the remediated soils are unlikely to cause adverse ecological effects. This finding is particularly important for post-industrial coastal environments where historical mercury use in chlor-alkali processes has often resulted in persistent contamination of soils and sediments [37].

Although only a limited number of soil cores were analysed, the results provide valuable supporting evidence that the remediation efforts at the former industrial site have significantly reduced the potential for mercury release to the surrounding environment. Moreover, these findings complement the analysis of mercury concentrations in marine organisms conducted in the present study, contributing to a more comprehensive assessment of the environmental legacy of the former industrial activity in the coastal area near Vlora.

Table 1 presents the descriptive statistics for mercury (Hg) concentrations in Patella caerulea. The results indicated that mercury concentrations in P. caerulea at all sampling stations were below the maximum allowable limit for food safety (0.5 µg/g wet mass). This situation suggests that prolonged mercury release from sediments has not significantly increased mercury accumulation in local biota. Consequently, they are safe for human consumption [38]. The relatively low mercury levels detected, even near previously contaminated sites, could be due to mercury converting to highly insoluble forms such as mercury sulfide (HgS), which has a solubility product constant of approximately 10⁻⁵³ [39]. This transformation reduces mercury's bioavailability and its potential entry into the food chain [14].

In all cases, prior to performing analyses of variance (ANOVA), the assumptions of normality and homogeneity of variances were evaluated using the Shapiro-Wilk and Levene tests, respectively. The results showed that the data did not significantly deviate from normality and that variances were homogeneous (p > 0.05). Analysis of variance for Patella caerulea was applied to determine whether significant differences exist in the mean mercury) concentrations among the selected sampling stations. The results are shown in Table 2. When the p-value is less than the 0.05 significance threshold, the null hypothesis is rejected, indicating that the observed differences in mean mercury concentrations are statistically significant. Although the mercury concentration at station S1 did not reach the critical significance level, it was notably higher compared to other stations. The mean mercury concentration at S1 was significantly different from that at stations S2, S3, S4, S5, and S6, with p-values of 0.0258, 0.0001, 0.0002, 0.0039, and 0.0003, respectively.

These results suggest that historical mercury deposition at station S1 may still contribute to moderately elevated mercury levels in Patella caerulea collected from this site. The highest mercury concentration was recorded at station S1 (0.084 µg/g), followed by station S2 (0.048 µg/g), which is situated near the Naval Military Academy - an area still affected by illegal urban wastewater discharges.

Interestingly, the lowest average mercury concentration (0.020 µg/g) was detected at station S3, rather than at reference sites such as Porto Palermo or the Karaburun Peninsula. It had been expected that the military base area would represent the cleanest site, due to its relatively greater distance from the former soda factory. However, this area may still be influenced by past and ongoing military-related activities, including mechanical operations and the construction or maintenance of small military vessels, which could contribute to local contamination. In addition, the sampling site near the Karaburun Peninsula is located close to a fish farming facility, which may represent an additional local source of organic matter and trace elements through aquaculture-related activities. In contrast, station S3 appears to be characterised by relatively cleaner seawater conditions, likely influenced by substantial inputs of freshwater from nearby mountainous catchments that may dilute contaminants before they reach the coastal zone. Moreover, this station is located close to the drinking water intake area supplying Vlora, which further supports the interpretation that this site is subject to comparatively lower levels of pollution.

The mercury concentration results for M. galloprovincialis are presented in Table 3. The use of bivalve organisms as bioindicators for assessing chemical pollution in estuarine environments began in the early 1970s. Notably, Goldberg (1975) introduced the "Mussel Watch" program, describing it as a method to evaluate ocean health [40].

Farrington et al. reflect on 40 years of the "Mussel Watch" program, reviewing its evolution, effectiveness, and scientific impact in monitoring coastal water contamination using bivalves [41]. In Albania, mussels were first employed as bioindicators in the early 1990s. Upon comparison of our data on Mytilus galloprovincialis with previously obtained results for the same species collected in Durrës, Vlorë, and Sarandë, no statistically significant differences were observed [30].

Similar to findings for Patella caerulea, analysis of variance (Table 4) revealed significant differences in mean Hg concentrations at station S1 compared to other stations. The highest mercury concentration was detected at station S1 (0.076 µg/g), while the lowest mercury concentration was observed at station S3 (0.010 µg/g). In contrast, station S5 showed no statistically significant differences in mercury (Hg) levels when compared with the other stations.

Overall, the data on mercury accumulation in both Patella caerulea and Mytilus galloprovincialis provide no evidence of mercury contamination in seafood from historical elemental mercury pollution in the area. This absence of significant contamination may be due to the mercury in sediments, which exists primarily in elemental or inorganic forms that are poorly bioavailable to marine organisms and thus unlikely to enter the food chain.

The analysed fish ranged in mass from 15 to 500 grams. No significant correlation was observed between fish mass and mercury concentration. Most species exhibited average mercury content below 0.06 µg/g. Table 5 shows the mercury concentrations and descriptive statistics for the fish.

But, among the analysed species, three predatory fish, Trigla lyra, Sphyraena sphyraena, and Squalus acanthias showed higher average mercury levels compared to other species, with mean content of 0.154, 0.175, and 0.172 µg/g mercury, respectively. Some individuals of these species exhibited even higher mercury levels, with maxima of 0.207, 0.394, and 0.210 µg/g mercury, respectively. Statistical analysis using analysis of variance confirmed a significant difference in Hg concentrations between the three predatory species, Trigla lyra, Sphyraena sphyraena, and Squalus acanthias and the other analysed fish species (Table 6). According to FDA/EPA fish consumption guidelines, most fish species analysed from the Albanian Coast fall under the “Best Choices” category (less than 0.15 µg/g mercury), which recommends consuming 2–3 servings per week [38].

However, Trigla lyra, Sphyraena sphyraena, and Squalus acanthias are classified in the “Good Choices” category, which limits consumption to one serving per week. Fish with average Hg concentrations exceeding 0.46 µg/g mercury are categorised as “Choices to Avoid” by Food and Drug Administration / Environmental Protection Agency (FDA/EPA) guidelines [38].

In addition, it has been reported that frequent consumption of large predatory fish, such as swordfish, tuna, and halibut, may significantly increase methylmercury intake, potentially exceeding the tolerable weekly intake (TWI) [42].

Based on the results and in reference to the new EU regulation, the average mercury concentrations in the analysed fish samples were below the recommended permissible limits. However, one Sphyraena sphyraena individual exceeded the maximum mercury limit established by the latest EU regulation [43]. In Albania, public awareness of the risks associated with methylmercury in seafood is low. This situation is particularly concerning for vulnerable groups such as women and children, who are more sensitive to mercury exposure. It is essential to provide information on the potential health risks of consuming seafood with elevated mercury levels.