Mercury Levels in Sediment, Water and Selected Organisms Collected in a Coastal Contaminated Environment: The Marano and Grado Lagoon (Northern Adriatic Sea, Italy)

Abstract

Mercury (Hg) is a global pollutant capable of bioaccumulates/biomagnifies along the trophic chain and posing concerns for organisms and humans. The historical mining in Idrija (NW Slovenia) and the more recent activity of a chlor-alkali plant (CAP) sited in Torviscosa (NE Italy) causes diffuse Hg contamination in the Marano and Grado Lagoon (MGL, northern Adriatic Sea, Italy). Despite the importance of fishing and aquaculture for local inhabitants, knowledge of the Hg content of MGL fish is still scarce and fragmentary. This paper reports the results obtained from the collection of sediments, water, and biota during the implementation of the WFD/2000/60/CE. The solid phase is characterised by high Hg concentrations (up to 7.4 mg kg⁻¹) with a net positive gradient moving eastward, but chemical speciation suggests the prevalence of cinnabar (not mobile) species. The scarce mobility of Hg is attested to by the low concentrations found in surface waters. Hg in fish often exceeds the limit set for commercialization (0.5 mg kg⁻¹ ww), especially in the Grado basin, but its content is variable depending on the size and habits of species. Although there was a significant linear relationship between THg content in sediment and tissues of Chelon auratus, the values of the biota sediment accumulation factor (BSAF), which were always less than one, suggest that the Hg bioavailable for transfer from sediment to biota is low. Additionally, the Target Hazard Quotient (THQ) calculated on C. auratus’s daily consumption showed that adverse effects on human health are out of the question at least for the Marano basin.

1. Introduction

Mercury (Hg) is a naturally occurring element (from 0.021 to 0.056 mg kg⁻¹ in the Earth’s crust) [1]. It is ubiquitous in the environment and presents in three main forms: elemental, inorganic, and organic, all characterised by variable toxic effects for mammals [2,3]. Hg is released into the environment from both natural and anthropogenic sources, but the latter prevails, and consequently, Hg is nowadays widespread at levels that exceed the thresholds set by environmental laws in the atmosphere, soils, riverine, and ocean systems [4,5,6,7]. Hg enters the aquatic environment through wet and dry atmospheric depositions, riverine runoff, erosion of contaminated soils, and direct discharge [8]. Once in the water column, Hg settles and sinks in sediment where it accumulates but can also be re-mobilised to the upper water column (both in dissolved and particulate form), especially in case of resuspension due to natural (i.e., tidal currents, wind waves, storm events, and wave-current interaction) and anthropogenic (i.e., dredging, trawling, and fisheries activity) factors [9]. Under favourable conditions, Hg can be methylated by specific microorganisms (e.g., sulfate-reducing bacteria, SRB, Fe(III)-reducing bacteria) to form methylmercury (MeHg), an organic species that, unlike the inorganic form, tends to bioaccumulate through the trophic chain [10] and be biomagnified up to final consumers [11]. The occurrence of neurotoxic and genotoxic effects in cases of both acute and chronic exposure to Hg and its compounds, especially MeHg, is well known [12,13]. To protect the environment and human health, the Minamata Convention on Hg was implemented, and it is actually signed by 140 countries worldwide [14].

Extraction of Hg from mining, chlor-alkali plants (CAP), coal mining, industrial wastewaters, gold mining, the chemical industry, munition manufacturing, artisanal small-scale gold mining, agriculture runoff, and oil refineries are the main anthropogenic activities responsible for the worldwide contamination of sediments [8]. In detail, mining activity is recognised as the main source of Hg for downstream sediments, with a continuous increase in extraction activity from 1500 to 1970, with peaks corresponding to silver mining in Spanish America, gold mining in North America, World War II, and massive industrial uses, followed by a steep decline until today [15]. The Mediterranean Sea can be considered a Hg hotspot for contamination of sediments, water, and biota [16]. In this area, the extraction of cinnabar (α-HgS) and native Hg (Hg⁰) was conducted starting from the Roman Age until the last decades of the 20th century [17,18,19], and a variable degree of contamination has been reported for air, water, and sediment [20].

The historical mining activity (about 500 years) conducted at Idrija (NW Slovenia) [21] caused diffuse Hg contamination in the sediments of the Isonzo/Soča River and of the alluvial plain [22], the Gulf of Trieste, and Marano and Grado Lagoon (hereafter referred to as MGL) [23,24,25,26]. Even though the mining district closed in 1995, it is clear that the 35,000 T of mercury that has been released into the environment [27] is still a problem. The MGL, which is the area investigated in this paper, is still receiving periodic inputs of particulate material enriched in Hg under favourable meteorological conditions (precipitation and ENE wind direction), which enter the lagoon through the easternmost tidal inlets (Primero; about 1.4 kg of Hg in a tidal semi-cycle) [28], whereas a partial loss of enriched Hg particulate material was calculated for the Grado inlet [29]. The water from the Aussa River also flows into the central part of the MGL, which is called the Buso basin. This latter drains an industrial site where a CAP operated from 1949 and discharged uncontrolled effluents until 1984. About 186 T of Hg were deposited in the LGM [30], and the inputs of contaminated particles and water are still considerable [31].

The first study on Hg in the MGL was conducted by Brambati [32], who reported a variable degree of contamination in sediment (from 1 to 10 mg kg⁻¹) and biota. Piani et al. [30] made the first effort to investigate the potential behaviour of Hg trapped in sediment by applying the pyrolytic speciation technique proposed by Biester et al. [33]. The authors found cinnabar compounds (mostly stable and insoluble), but only a partial area of the MGL was considered. A more comprehensive work was conducted in the framework of the MIRACLE (Mercury Interdisciplinary Research project for Appropriate Clam farming in Lagoon Environment) Project (2008–2009). The project aimed to understand Hg biogeochemical cycling by testing the coexistence of clam farming with Hg contamination in the sediments. Hg was measured in water, sediment, and biota, and its mobility and speciation were also considered together for bioaccumulation in clams [34]. The Hg distribution showed a positive gradient moving eastward (from 0.68 to 9.95 mg kg⁻¹), with MeHg accounting for 0.08% of the total pool on average [22] and a sediment thickness involved in the contamination of more than 1 m [35]. The potential mobility of Hg at the sediment-water interface was assessed by calculation of diffusive fluxes, in situ deployment of benthic chambers (natural conditions) [36,37], and mesocosm experiments (perturbed conditions) [9]. The authors stated that sediments could be a primary source of MeHg for the water column, especially in cases of anoxic events, and that this transfer is more active in the Grado sector, whereas the effects of a resuspension are limited in time and trivial if compared to lagoon-wide processes. Subsequent research supports the hypothesis that unusual habitats, such as salt marshes, actively participate in Hg recycling [38,39]. It is well-known that the determination of total Hg is not useful to understanding the fate of this element and its potential toxicity for the aquatic environment. Thus, numerous speciation techniques are currently applied to define its behaviour [40]. In the MGL, both chemical and thermo-desorption techniques were applied and revealed the presence of both cinnabar and non-cinnabar compounds in variable percentages [26,30].

Molluscs and fish play an important role in the MGL’s economy [41] and seafood consumption is extremely beneficial for the population as a source of proteins, fatty acids, vitamins, and minerals. However, in a contaminated environment such as the MGL, the risks associated with the exposure of Hg/MeHg cannot be neglected and should be carefully considered [42]. In the work conducted by Brambati, the degree of Hg bioaccumulation was quite variable, with the highest values (up to 5 mg kg⁻¹ wet weight, ww) commonly found for carnivorous fish [32], whereas in the soft tissues of Manila clams (Ruditapes philippinarum), Hg contents were significantly lower (0.051–0.9 mg kg⁻¹ ww) [43,44]. Recently, Acquavita and Bettoso [45] investigated the relationship between total Hg content and some specific factors (i.e., sampling sites and season, size, and sex) for the grass goby Zosterisessor ophiocephalus. Total Hg was 0.61 ± 0.28 mg kg⁻¹ (ww), on average, with a maximum of 1.67 mg kg⁻¹ ww. Thus, the strict legal limits established for marketed sea foods [46] and for water quality classification [47] (0.5 and 0.02 mg kg⁻¹ ww, respectively) are often overcome.

In this paper, we report the results obtained from a complete survey on sediments, water, and biota conducted during the implementation of the WFD/2000/60/CE (from 2010 to 2021) [47]. The sediments were analysed for total Hg contents and speciation to determine potential bio-accessibility for biota [48]. Moreover, the quality of sediment was checked through the application of common indices of contamination [49,50,51]. Data on total Hg in the water column were scarce and referred to a limited area of the MGL [39], thus the results reported in this paper represent a novelty. Biota were collected with species from the Estuarine Usage Functional Group (EUFG) categories in mind [52]. The compliance with the imposed limits was verified for all species selected, and due to its habits, Chelon auratus (golden grey mullet) was selected to calculate the biota sediment accumulation factor (BSAF) [53] and the target hazard quotient (THQ) [54].

2. Materials and Methods

2.1. Study Area

The MGL (northern Adriatic Sea, Italy), which is located between the Tagliamento and Isonzo River deltas, covers an area of about 160 km² and, based on morphologic and hydrologic features, displays six main sub-basins. The MGL has been protected by the Ramsar Convention [55] due to its ecological importance since 1971 and is also designated as a Site of Community Importance (SCIs–IT3320037). On the other hand, numerous economic activities (i.e., tourism, fishing, and aquaculture) and the presence of the industrial hub sited inland in the Friulian Plain raise environmental concerns. Due to this, part of the lagoon was declared, in the past, a polluted site of national interest (SIN) [41,56].

The Isonzo and Tagliamento Rivers runoffs primarily transport solid suspended matter into the MGL, and the balance between sediment loss and accumulation is influenced by both natural and anthropogenic factors [57,58]. Freshwater inputs are almost confined in the western sector (Marano), due to the Stella and Cormor Rivers (flow rates of 36.1 and 10.7 m³ s⁻¹, respectively), whereas the contribution of the Turgnano, Zellina, Aussa-Corno, Natissa, and Tiel Rivers and the numerous drainage pumps located inland is less important. Marine waters enter the lagoon, moving northward, via the tidal inlets of Grado, Porto Buso, and Lignano. Semidiurnal tides ensure water exchange and cause significant daily salinity variability (more than 10 PSU). Overall, the lagoon exhibits a significant positive salinity gradient moving both outward and eastward [59]. The same authors reported significant spatial and seasonal variability for pH and dissolved oxygen. The freshwater inputs influence the nutrient distribution (nitrates are significantly enriched in the Marano sector) [60] and, as a consequence, the trophic state of the system [61,62].

2.2. Sampling and Analysis

Starting in 2010, the Environmental Protection Agency of the Friuli Venezia Giulia Region (ARPA FVG) planned numerous monitoring activities in the MGL according to the Water Framework Directive [47]. The final goal was to classify the quality of biological elements in seventeen water bodies (hereafter referred to as WBs) that were defined following the Italian Legislative Decrees [63] (Figure 1).

Surface sediments (n = 17) were collected using a van Veen grab (A = 0.047 m²), transferred into pre-cleaned bottles, frozen (T = −25 °C), and, before analysis, air-dried, sieved, and homogenised. Total Hg (THg) was determined with a Direct Mercury Analyser (DMA-80, Milestone, Sorisole (BG), Italy) following the EPA method 7473 [64]. The limit of quantitation (LOQ) was <0.06 mg kg⁻¹. The accuracy was checked via periodic proficiency tests conducted on certified samples (UNICHIM). The speciation of Hg was conducted following the selective sequential extraction (SSE) proposed by Bloom et al. [48]. Additionally, the obtained fractions were analysed with atomic fluorescence spectrometry (EPA method 1631e, CVAFS, Brooks Rand MERX-T, Seattle, WA, USA; LOQ = 4.0 ng L⁻¹) [65]. The recovery percentage was checked using reference standard materials (NIST-1646a = 0.04 mg kg⁻¹ and BCR-580 = 132 ± 3 mg kg⁻¹; 77.9 and 104%, respectively) as well as real samples (95 ± 19%).

Surface waters were collected with a telescopic bar coupled to pre-cleaned borosilicate bottles, filtered through 0.45 µm pore-size membrane filters, and added to 5 mL L⁻¹ of BrCl solution [65]. The total dissolved Hg (THg_D) was finally determined with the same instrumentation and method employed for Bloom’s fraction.

Fish samples of numerous species were initially collected for total Hg (THg) analysis during 2010 and 2011, simultaneously with the monitoring of the fish fauna for the biological quality element (BQE) according to [47]. During spring 2012, targeted monitoring of specimens of the big-scale sand smelt A. boyeri was conducted at twenty-two sampling stations. For each station, 100 individuals were randomly selected and 40 g of fish were homogenised for THg analysis. Finally, the golden grey mullet, C. auratus, was selected as a target species for monitoring priority pollutants in biota sensu WFD. Thus, a total of three specimens in each WB were caught during 2021. The total length (TL, cm) was measured in all specimens, and fish fillets for THg analysis were selected from the largest individuals, whereas juveniles and species characterised by small size (e.g., A. boyeri) and mullets were analysed in toto. All catches were conducted using the most appropriate and sustainable method depending on the targeted species. THg analyses were conducted by EPA 7473 [64] directly on fresh samples. Quality assurance was performed by blank control, analysis of replicates, and the use of certified reference material (DORM-2).

Statistical analysis was performed using EXCEL™ and PAST (Paleontological Statistics, Version 4.08) statistical programmes [66]. The non-parametric Kruskal-Wallis H test assesses significant differences between two or more groups of an independent variable [67]. The Mann-Kendall trend test was employed to test for the presence of significant temporal increase/decrease in the selected parameters [68]. The distribution maps were realised using the QGis Version 3.16.3 “Hannover” program.

3. Results

3.1. Total Hg and Speciation in Sediment

The results obtained in surface sediments are summarised in

Table 1. Total Hg concentration, chemical fractionation, and calculated I_geo [49] in surface sediments of the Marano and Grado Lagoon (n.d. = not determined).

Water Body	Total Hg (mg kg−1)	F1 (%)	F2 (%)	F3 (%)	F4 (%)	F5 (%)	Igeo	Sediment Quality	Class
FM1	7.38	0.39	0.04	6.85	21.65	71.08	14.8	EC	6
FM2	5.67	0.26	0.02	4.71	40.23	54.78	12.8	EC	6
FM3	6.20	n.d.	n.d.	n.d.	n.d.	n.d.	13.5	EC	6
FM4	5.09	0.18	0.03	7.28	55.21	37.30	12.0	EC	6
TEU1	3.96	0.17	0.02	20.52	15.30	63.99	10.2	EC	6
TEU2	4.42	0.22	0.08	22.28	23.57	53.86	11.0	EC	6
TEU3	5.55	0.14	0.03	8.75	30.28	60.80	12.7	EC	6
TEU4	6.66	0.12	0.01	14.76	15.03	70.07	14.0	EC	6
TME1	2.77	0.30	0.05	20.21	18.07	61.36	7.5	EC	6
TME2	1.73	0.36	0.04	10.61	33.69	55.30	4.1	H to EC	5
TME3	1.91	0.21	0.04	10.87	29.81	59.06	4.8	H to EC	5
TME4	1.52	0.35	0.04	3.10	24.09	72.41	3.1	HC	6
TPO1	5.21	0.15	0.05	4.72	67.24	27.46	12.2	EC	6
TPO2	3.75	0.15	0.02	25.02	18.60	56.22	9.8	EC	6
TPO3	2.18	0.45	0.13	7.30	46.76	45.35	5.8	EC	6
TPO4	1.66	0.53	0.04	6.31	23.94	69.18	3.7	HC	4
TPO5	1.30	0.52	0.07	8.64	35.64	55.13	1.9	MC	2

Note: EC = extremely contaminated; H to EC = heavily to extremely contaminated; HC = heavily contaminated; MC = moderately contaminated.

.

Total Hg was 3.94 ± 2.01 mg kg⁻¹ (mean ± s.d.) and displayed a high degree of variability ranging from 1.30 to 7.35 mg kg⁻¹ (at TPO5 and FM1, respectively). The spatial distribution showed a clear positive gradient moving eastward (from Marano to Grado basin) and also a significant enrichment in the central sector (TEU4 = 6.66 mg kg⁻¹) (Figure 2). These results confirm the evidence reported by Acquavita et al. [26] (mean = 4.62 mg kg⁻¹), where a comprehensive historical dataset (n = 178) was mapped, and, overall, the results of previous studies conducted at different sites of the same lagoon [26,30,36,39,69,70,71]. It should be noted that these results were two orders of magnitude higher than the Hg natural background (0.13 ± 0.04 mg kg⁻¹) calculated by investigation in deep cores (n = 13; more than 1 m of depth) collected throughout the basin [35].

The threshold set for the good quality of sediment sensu WFD, which is 0.3 mg kg⁻¹ (deviation permitted = +20%), was exceeded for all the sampling sites. This was also true for the derived effect range low and median (ERL and ERM; 0.15 and 0.71 mg kg⁻¹, respectively), previously calculated by Long et al. [50], which have been shown to successfully predict the occurrence of potential toxicity [51]. The calculated geoaccumulation index (I_geo = log₂ C_n/1.5 × B_n, with C_n as the concentration of the element considered, B_n as natural background, and 1.5 as a factor to account for variations due to lithogenic effects), introduced and scaled by Müller [49]. This revealed a high degree of sediment contamination (I_geo = 1.9–14.8; on average, 9.1 ± 4.3), with 76% of sites belonging to class six (extremely contaminated), indicating at least a 100-fold enrichment factor above background values (I_geo > 6) (Table 1).

Despite the natural sources present in the Mediterranean area, mining and anthropogenic activities are responsible for numerous Hg hotspots in coastal sediments, and the MGL with 271 Mg of buried Hg is one of the most polluted [72]. The nearby Venice Lagoon [73,74,75,76,77] and the Santa Gilla Lagoon (Sardinia, Italy), both impacted by CAP plants [78], showed comparable degrees of contamination, whereas the Piallassa Baiona (acetaldehyde plant) and the Orbetello Lagoon (groundwater discharge from mining at Monte Amiata) were up to 1 or 2 orders of magnitude more contaminated [70,79,80,81,82,83,84,85,86]. On the other hand, it should be emphasized that other environments display values often equal to the background (see

Table 2. Overview of total Hg in the sediments of some transitional environments in the Mediterranean Sea.

Site	Total Hg (mg kg−1)	Reference
Piallassa Baiona (Italy)	n.a.–160	[79]
Piallassa Baiona (Italy)	0.13–250	[81]
Piallassa Baiona (Italy)	11–43	[82]
Piallassa Baiona (Italy)	0.88–38	[80]
Piallassa Baiona (Italy)	0.37–5.51	[83]
Piallassa Baiona (Italy)	0.72 ± 0.3–22.79 ± 6.7	[85]
Piallassa Baiona (Italy)	14.4–79.0	[72]
Santa Gilla (Sardinia, Italy)	0.206–8.63	[80]
Varano Lagoon (Italy)	0.04–0.04	[87]
Lesina (Italy)	0.04–0.12	[88]
Stagnone Marsala (Italy)	0.18 ± 0.01–0.67 ± 0.06	[89]
Orbetello Lagoon (Italy)	0.30–2.64	[88]
Orbetello Lagoon (Italy)	0.57–37.63	[84]
Orbetello Lagoon (Italy)	0.56–28.18	[86]
Venice Lagoon (Italy)	0.05–3.8	[73]
Venice Lagoon (Italy)	0.1–1.9	[74]
Venice Lagoon (Italy)	0.64–3.41	[75]
Venice Lagoon (Italy)	0.1–3.4	[76]
Venice Lagoon (Italy)	0.03–3.9	[77]
Berre Lagoon (France)	0.15–0.40	[90]
Berre Lagoon (France)	0.068 ± 0.002–0.725 ± 0.019	[91]
Bizerte Lagoon (Tunisia)	0.008–0.64	[92]
Bizerte Lagoon (Tunisia)	0.007 ± 0.002–0.102 ± 0.004	[93]
Marano and Grado Lagoon (Italy)	1.62–10.06	[69]
Marano and Grado Lagoon (Italy)	4.1–6.6	[30]
Marano and Grado Lagoon (Italy)	9.5 ± 0.2–14.4 ± 0.7	[36]
Marano and Grado Lagoon (Italy)	10.7–12.5	[70]
Marano and Grado Lagoon (Italy)	0.68–9.95	[26]
Marano and Grado Lagoon (Italy)	2.15–6.87	[71]
Marano and Grado Lagoon (Italy)	3.79–7.25	[39]
Marano and Grado Lagoon (Italy)	1.3–7.38	This work

for a comprehensive overview).

The fate of Hg accumulated in sediment is not fully predictable because only specific chemical forms have the potential to be methylated and bioaccumulated/biomagnified through the trophic chain [40,94]. In our case, to elucidate both the potential mobility and bioavailability, we select Bloom’s procedure, which can be summarised in terms of classification and extracted compounds as reported in

Table 3. Primary compounds are released with Bloom’s selective sequential extraction [49].

Fractions	Hg Classification	Primary Compounds Extracted
F1	Water-soluble, i.e., salts	HgCl2
F2	Weak acid-soluble/”stomach acid” soluble	HgSO4, HgO
F3	Organo-complexed	Hg-humics, Hg2Cl2, MeHg
F4	Strongly-complexed	mineral lattice bound, Hg2Cl2, Hg0
F5	Mineral-bound	HgS, m-HgS, HgSe, HgAu

.

The fractions F1 and F2 represent the readily soluble species and can be attributed to the bioaccessible form of Hg (i.e., HgCl₂, HgSO₄, and HgO) in the mammalian gastrointestinal tract, thereby being commonly associated with the most dangerous fractions. In this work, F1 and F2 represent the lowest content with respect to the total Hg bulk, with percentages of 0.28 ± 0.13 and 0.04 ± 0.03%, respectively (Figure 3). Overall, the maximum value of the F1 + F2 sum was 0.59% (mean ± std. dev. 0.33 ± 0.16%), which can also be interpreted as the potential limit for inorganic Hg available for methylation processes.

These findings are comparable to those reported for sediment from the Gulf of Trieste, the mouth of the Isonzo River, Piallassa Baiona, and Taranto Harbour, but significantly lower than those found in the Aussa-Corno River (see

Table 4. Speciation of Hg in the sediment of some coastal sites obtained by Bloom speciation. Modified from Covelli et al. [95].

Site	Hg (mg kg−1)	F1 (%)	F2 (%)	F3 (%)	F4 (%)	F5 (%)	References
Aussa River	0.82–5.69	0.66–5.60	0.96–6.05	5.59–24.81	33.11–90.57	0–57.95	[31]
Grado Lagoon	10.75–13.37	0.16–0.20	0.01–0.02	2.72–4.28	43.58–53.31	42.24–53.49	[70]
Pialassa Baiona	14.40–19.10	0.20–0.38	0–0.02	4.77–6.32	87.00–87.81	5.67–7.83	[70]
Gulf of Trieste	6.36–13.50	0.08–0.18	0.04–0.18	0.71–1.47	32.6–59.7	39.3–56.7	[95]
Taranto harbour	6.61	0.30	0.00	4.20	93.10	2.30	[96]
Isonzo River mouth	13.27	0.08	0.03	0.77	23.45	75.68	[97]
Marano Grado Lagoon	1.30–7.38	0.15–0.53	0.01–0.13	3.1–25.0	15.3–67.6	27.5–72.4	This study

modified from Covelli et al. [95]).

The percentages obtained for fraction F3 were quite higher (11.4 ± 7.0%) and showed a high degree of variability (from 3.1 to 25.0% at TME4 and TPO2, respectively). In detail, TEU1, TEU2, and TPO2, which are sites belonging to the easternmost sector of the lagoon, and TME1, close to the Aussa-Corno and Zellina River mouths, were characterised by percentages up to 20%, probably as a consequence of their enrichment in an organic matter [22]. The F3 fraction includes the organo-chelated forms of Hg (i.e., humic, fulvic, and amino acids, living and dead biota) that have moderate mobility. To a lesser extent, this fraction can be correlated to the presence of MeHg [48]. However, as previously stated by Acquavita et al. [22], this toxic form in the MGL accounted, on average, for 0.08% of the total Hg pool. This suggests that in the sediments of the MGL, Hg with scarce mobility predominates. F4 includes strongly complexed Hg forms in industrial products such as calomelan (Hg₂Cl₂). This was also suggested by Bloom et al. [48] as an estimate of the presence of free elemental Hg, which commonly originates from industrial plant discharges (i.e., CAP) [31,70,98,99]. However, in this work, it is more reasonable to assume that the high percentage found (31.2 ± 15.0%) is representative of Hg bound up in amorphous organo-sulphur or crystalline Fe/Mn oxide phases, which are strongly bound species. The F5 (57.1 ± 12.3%), was the most abundant fraction, representing the more insoluble species (Kps~10⁻⁵³) [100] such as cinnabar (α-HgS) and meta-cinnabar (β-HgS). This is consistent with the hypothesis that the main input of Hg in the form of particulate matter can be attributed to the Isonzo River discharge influenced by the contamination of both soils and sediments in the plain of the former mining area of Idrija [22,101].

As aforementioned, the percentages of the fractions extracted differed significantly (K-W: p_(same) = 4.99 e⁻¹⁵). However, a significant correlation was found for F1/F2 (r = 0.05329), F3/F4, and F4/F5 (r = 0.0667 and 0.08975, respectively). Unfortunately, there were no data on secondary parameters such as grain size and organic matter (C_org) to check the occurrence of significant correlations with the single fractions detected.

3.2. Total Dissolved Hg in Surface Waters

Total dissolved Hg (THg_D) was determined seasonally from November 2017 to November 2019, for a total of eight campaigns. The concentrations were quite variable, ranging from 1.6 to 28.7 ng L⁻¹ (mean 6.6 ± 5.4 ng L⁻¹), with no values exceeding the limit set by the WFD (SQA-CMA = 70 ng L⁻¹). On average, FM1 and TEU 3 showed the highest values with 11.4 ± 8.4 and 11.1 ± 10.6 ng L⁻¹, respectively, whereas the lowest value was recorded at TPO5 (4.4 ± 2.0 ng L⁻¹). However, no significant difference was found among the WBs (K-W: p_(same) = 0.9517) (Figure 4). This is probably due to the high hydrodynamics of the system, which is characterised by a low residence time of water (1.9 ± 1.6 days) calculated by Ferrarin et al. for the whole basin [102].

Taking into account the temporal trend, it should be noted that there was a significant decrease in THg_D concentrations at TEU2, TME2, TPO3, and TPO5. These results are significantly lower than those found in some sites of the nearby Gulf of Trieste (up to ~50 ng L⁻¹ in surface waters of the Panzano Bay, Pavoni et al. [103]). Additionally, these results are more comparable to those reported for the Isonzo River mouth, Timavo River estuary, and other sites in the Gulf of Trieste [28,104,105,106]. Horvat et al. [107] reported values significantly lower at Kaštela Bay (contamination due to CAP), where the highest value reported was 0.54 ng L⁻¹.

3.3. Total Hg in Biota

Lagoons are commonly used as temporary habitats by fish in terms of feeding and/or nursery grounds [52]. The marine components are the dominant contributors to the diversity of transitional water fish fauna [108], and, in turn, relatively few species use these environments for spawning or permanent residence, as well as fewer use estuarine/lagoon systems for diadromous migrations [52]. Based on this assumption, all the fish species analysed in this work were classified following the Estuarine Usage Functional Group (EUFG) categories proposed by Franco et al. [52] and the summary of the results is reported in

Table 5. Total Hg concentrations in selected fishes collected in the Marano and Grado Lagoons.

EUFG	Species	Year	Ltot (cm)	THg (mg kg−1)	Ltot (cm)	THg (mg kg−1)	Ltot (cm)	THg (mg kg−1)
Estuarine	Atherina boyeri	2012	7.8–8.6	0.17–0.28	7.6–8.1	0.23–0.78	8.4	0.72
	Salaria pavo	2010/2011	8.2	1.74	8.3	1.6	nd	nd
	Zosterisessor ophiocephalus	2010/2011	7.4–12.6	0.16–0.49	nd	nd	10.2	0.61
	Aphanius fasciatus	2010/2011	4.1	0.21	nd	nd	4.4–4.8	0.4–0.74
	Pomatoschistus canestrinii	2010/2011	4.4	0.04	nd	nd	nd	nd
Diadromous	Alosa fallax	2010/2011	28.5	0.76	nd	nd	nd	nd
Marine Migrant	Engraulis encrasicolus	2010/2011	9	0.07–0.17	nd	nd	nd	nd
	Chelon auratus	2021	24.6–30.9	0.1–0.18	24.3–29.5	0.23–0.5	21	0.39

Note: in bold are the results on individuals collected in the Grado Lagoon, and in italics, are those collected at Val Cavanata (FM1). n.d.: not determined.

.

THg in the collected fish fauna ranged from 0.04 mg kg⁻¹ ww found in the Canestrini’s goby (Pomatoschistus canestrinii) to 1.74 mg kg⁻¹ ww detected for the peacock blenny (Salaria pavo). It is clear that, beyond the EUFG assigned to every fish species, the individuals collected in the Grado basin (especially at Val Cavanata, FM1) were generally characterised by higher Hg concentrations than those collected at Marano. This confirms the evidence reported by Brambati [69], in which an increasing Hg gradient in fish was observed moving eastward. In addition, the further degree of bioaccumulation found at FM1 could derive from some peculiar characteristics of this area. It is a disused fish farm connected to the open sea through a unique sluice gate. Therefore, it is a very confined environment that suffers from very scarce water renewal. Here, the biogeochemical processes are very intense, with the occurrence of ipoxic/anoxic events, especially in late spring and summer [61], that can favour MeHg production, as observed in similar environments [36], and the subsequent bioaccumulation and biomagnification of Hg.

Among the estuarine species, the big-scale sand smelt (A. boyeri) is one of the most abundant resident fish of Mediterranean coastal lagoons [109], where it represents a key role for the estuarine food web, being one of the most important links between primary benthic and planktonic consumers and the higher trophic levels [110]. A. boyeri is not a sedentary benthic species, but it is characterised by a certain degree of site fidelity, in particular to spawning grounds, thus resulting in semi-isolated populations [111]. This species represents one of the most important fisheries resources in the MGL, being caught by fyke nets, a traditional fishing gear widespread in the northern Adriatic lagoons [112]. The TL of sand smelts collected at the twenty-two sampling sites ranged from 6.26 ± 0.86 to 8.42 ± 0.79 cm (mean value 7.29 cm), whereas THg ranged from 0.28 to 0.85 mg kg⁻¹ ww (mean value 0.52 ± 0.18 mg kg⁻¹ ww).

Regarding the solely estuarine species, which are confined to estuarine habitats and complete their entire life cycle within these environments, the Canestrini’s goby (P. canestrinii) is an endemic species of the Adriatic lagoon and estuarine areas, where salinity ranges between 5–20, and the sedimentary habitat is without vegetation [113,114]. Moreover, it is included in Annex II of the Habitat Directive [115]. This species was recorded only in the most confined areas of the Marano basin close to the river mouths [112], and the THg data obtained in this work represents a novelty for adult male specimens of this species.

The Mediterranean banded killifish (Aphanius fasciatus) is a typical euryhaline fish species that inhabit the brackish environment, lagoons, estuaries, and salina. It is rarely found in freshwater habitats [116]. It is also included in Annex II of the Habitat Directive, and, due to its euryhaline characteristics, it was constantly recorded in every WB type of the MGL, being particularly abundant at FM1 [112]. THg concentrations were significantly higher in FM1 specimens than in Marano specimens. Unfortunately, specimens from Grado were not collected, but it should be noted that, based on data published by Brambati [69], individuals of 4–4.5 cm TL caught in a fish farm in the eastern area of the Grado Lagoon showed THg values ranging from 0.49 to 1.7 mg kg⁻¹.

The peacock blenny (S. pavo) is typical in the intertidal zone and very shallow water (from 0 to 2 m), and it is a euryhaline species commonly found in brackish waters such as lagoons [117]. Only two adult specimens were analysed and both of them revealed the highest THg values recorded (1.6 and 1.7 mg kg⁻¹ ww).

The grass goby (Z. ophiocephalus) is a brackish goby widely distributed in the estuarine and lagoon environments of the Mediterranean basin [118,119] and is particularly abundant in the northern Adriatic Sea, Black Sea, and the Sea of Azov [120]. In the largest northern Adriatic lagoons (Venice and Marano—Grado Lagoon), where its exploitation has occurred since the twelfth century, it is one of the most important target species of artisanal fisheries [119]. The sexual maturity of Z. ophiocephalus occurs in the first year of life when individuals reach a size of approximately 7.5 cm in TL [119]. The TL of the specimens examined in this study ranged from 7.4 to 12.6 cm, and, as with other species, the highest values of THg peaked at FM1 (0.61 mg kg⁻¹ ww), whereas the minimum value (0.16 mg kg⁻¹ ww) was recorded in the smaller individual (7.4 cm TL) caught in the Marano basin. Nevertheless, the general assumption of the occurrence of a significant positive relationship between TL and THg, at least for this species, cannot be assumed. Acquavita and Bettoso [45] investigated the THg content within a population of 208 individuals caught in the Marano Lagoon and found the highest levels in females (0.74 mg kg⁻¹ ww) rather than males (0.48 mg kg⁻¹ ww), even though the captured males were bigger than females (18.5 and 16.0 TL cm on average). The authors hypothesised that THg bioconcentration for this species is affected not only by individual size and environmental characteristics of the sampling area but also by gender and reproductive season.

The twaite shad, Alosa fallax, is an anadromous species that spends most of its life at sea and migrates to freshwater environments for breeding [121]. Only one specimen was successfully caught in this study (Marano basin; 28.5 cm in TL) and 0.76 mg kg⁻¹ ww of THg were measured in the muscle tissue, which is approximately three times higher than the median value recorded in adult specimens collected from the Gironde Estuary (South West France) [122]. This further confirms that the THg concentrations of fish in the Mediterranean Sea are higher compared to those caught in the adjacent Atlantic Ocean [123].

The anchovy, Engraulis encrasicolus, is a marine migrant and planktonic feeder that uses the lagoon habitat as a nursery ground. Only two specimens of 9 cm TL were analysed for Marano, with a THg concentration of 0.07 and 0.17 mg kg⁻¹ ww. These values fall within the range found in specimens (n = 45) from the eastern Adriatic (Croatia) (0.001–0.52 mg kg⁻¹; mean value 0.04 mg kg⁻¹ ww) [124], whereas the average value was 0.05 and 0.028 mg kg⁻¹ ww for the Mediterranean and Atlantic Oceans, respectively [123,125].

The golden grey mullet C. auratus was selected as the target species for priority pollutants monitoring in biota sensu WFD because it is one of the dominant species in the surveyed lagoon, being easy to identify and capture. It is a marine migrant species with a detritivore feeding strategy [52], and its life history makes it particularly appropriate to the current goals [126]. The TL content of FM1 specimens was slightly lower than that of the Marano and Grado basins, but the THg content was at least twice as high in FM1 and Grado as it was in Marano. The significant linear relationship (r = 0.7249; p < 0.001) occurring among these variables suggests the influence of THg content in sediment on bioaccumulation for this detritivore species and confirmed the observation reported in the Ria de Aveiro (Portugal), a lagoon adjacent to the Atlantic Ocean that received Hg-containing effluents from a CAP from the 1950s until 1994 [126].

Taking into consideration that this species lives in proximity (contact) to the sediment, it is suitable for the calculation of the biota sediment accumulation factor (BSAF) [53] according to the formula:

BSAF = C_(biota)/C_(sediment)

The calculated values ranged from 0.023 to 0.098 (TEU4 and FM4, respectively), which were always lower than 1 and suggest that the Hg bioavailable for the transfer from sediment to biota is small (Figure 5).

Mancini et al. [127] reported significantly higher values (up to 0.2) calculated for European seabream (Sparus aurata) and seabass (Dicentrarchus labrax) collected in the Orbetello Lagoon (Tuscany). However, it should be noted that these species have different feeding habits with respect to grey mullet and are at an upper level along the trophic chain. Taking into consideration the results found for Hg speciation, no significant relationships were found between the calculated BSAFs and the fractions investigated.

The consumption of fish accounts for about 17% of the total proteins provided to the population with the diet [128], and it is strongly recommended also for the assumption of minerals, vitamins, and omega-3 long-chain polyunsaturated fatty acids (LC-PUFA) [129]. Therefore, the calculation of the potential exposure/risk to humans must be carefully considered. In this work, we considered the target hazard quotient (THQ) [54,130,131], which is calculated as follows:

THQ = [(EF × ED × FIR × C/RfD × BW × AT)] × 10⁻³

where EF is exposure frequency (365 days/year), ED is exposure duration (70 years), equivalent to the average lifetime, FIR is food ingestion rate (in Italy: 77.8 g day⁻¹) [132], C is THg concentration in fish (mg kg⁻¹ ww), RfD is oral reference dose (Hg = 3.0 × 10⁻⁴ g day⁻¹), BW is body weight (60 kg), and AT is the averaging exposure time for non-carcinogens (365 days/year × ED). If the THQ value obtained is below “1,” an adverse effect is out of the question in terms of human health. In this case, the selected target species was C. auratus.

THQ showed a wide range of variability, ranging from 0.45 to 2.16 (at TPO4 and FM4, respectively). In detail, the WBs located in the Marano basin always displayed values < 1, whereas in the Grado basin are generally >1 (Figure 6).

4. Discussion

A complete investigation of Hg contamination in the MGL is an issue that cannot be neglected because of the important ecological and socio-economic values of this coastal ecosystem [41]. As previously stated, the contamination of surface sediment involves the whole lagoon area and up to 1 m in depth [22,35]. Thus, it represents a permanent issue, and in toto reclamation is unfeasible for both economic and ecological concerns. The effects of re-suspended sediments on Hg biogeochemical cycling are well known and still a matter of debate [9,83,132,133,134]. The situation is comparable to that found inland through the Isonzo River plain, where specific risk management methods are currently under investigation [22]. The MGL is also an important habitat for fish fauna and a nursery ground for juvenile and migratory species, and it has been a suitable location for fish farm activities since the 16th century. Nowadays, about forty fish farms of variable dimensions and potential productivity are still active [112], and the restoration of some abandoned areas is in progress.

Today the main source of Hg and its more toxic form, MeHg, for humans is fish consumption [135] and exposure can cause progressive damage to the nervous system, particularly in foetuses and children (i.e., decreasing fine motor-adaptive function, reduction in IQ, attention deficit disorder) [136]. It is well known that the degree of bioaccumulation/biomagnification along the trophic chain depends on several factors. At first glance, total Hg content appears to be the most important factor; however, MeHg production via specific bacterial activity and, to a lesser extent, abiotic processes are strongly dependent on key factors such as physico-chemical conditions (i.e., oxygen content, pH, solar radiation/temperature, labile natural organic matter, water residence time), and almost entirely on Hg speciation [100,137,138]. Thus, it surprises that heavily contaminated sites (i.e., Pialassa Baiona) do not show a correspondent bioaccumulation in selected organisms [81]. In the MGL, the not-mobile fraction of Hg (Hg-S) prevails, and this is a factor that normally limits the bioavailability of Hg for methylation. On the other hand, the most contaminated areas of the Grado basin are characterised by a higher degree of bioaccumulation. In addition, relative production/fluxes of MeHg seem to be site-specific, being quite different among sites characterised by comparable Hg contamination [26,36]. As a result, it is possible to hypothesise that a better understanding of organic matter composition and its behavior, whether labile or not, could be useful in determining the potential for methylation [137]. It should be noted, however, that the LMG shows comparable Hg recycling coupled to MeHg production in the adjacent Gulf of Trieste and that this is more intense during the summer period [36]. It can be noticed that, despite the Hg contamination, the relative amount of Hg in the water column does not exceed the limit set by the WFD, and it was significantly lower than those observed in the Gulf of Trieste, probably as a consequence of the solid-phase porewater distribution coefficient (K_D, L kg⁻¹) [139] that is about one order of magnitude (10⁶ vs. 10⁵) higher than that of the gulf.

Exposure to Hg (MeHg) through fish consumption is a great concern, especially during the pre-natal period [140], but the studies conducted in different contexts showed no clear evidence [141]. Due to the potential adverse effect on residents, epidemiological studies were also conducted on residents of the Friuli-Venezia Giulia region. Hg analysis in human hair from residents of the Gulf of Trieste revealed that a risk alert seems inappropriate, although it appeared prudent to limit the ingestion of local fish to <4 servings/month for pregnant women and children [142]. Based on the THQ estimation, a THg value of 0.23 mg kg⁻¹ ww in edible parts seems acceptable to avoid adverse effects on human health in the long term. In a transitional environment such as MGL, the THg concentrations in fish are positively correlated to those in the sediment, but bioaccumulation is also a trait characteristic for every species in relation to feeding guilds, migration, time of reproduction, gender, size, and so on. This means that any generalisation about THg bioaccumulation in marine organisms is pure conjecture. Risk analysis for fish consumption in areas impacted by Hg, such as MGL and the Gulf of Trieste, should consider a wide range of fish, molluscs, and crustacean species, beyond an acceptable sample number of specimens for every species. Furthermore, the selenium (Se) concentration in edible parts should also be analysed to estimate the degree of protection. Nevertheless, the THg value in Adriatic exceeded the limit of 0.5 mg kg⁻¹ ww in many fish species for commercialization in UE [46]. Notably, the calculated selenium/mercury (Se/Hg) molar ratio and the Se—Health Benefit Value (Se-HBV) ([143,144,145,146,147] for further details and application) in all fish analysed in the eastern Adriatic by Žvab Rožič et al. [148] showed that Se has positive influences on Hg detoxification. This indicates that Se excess occurs after Hg sequestration, and consequently, the consumption of these fish in human nutrition is not risky [148]. This was also confirmed by Sulimanec Grgec et al. [149], who analysed THg and Se in the muscle tissue of twelve commercially important fish species (n = 717) from forty-eight locations in the eastern Adriatic Sea (Croatia). Following these considerations, some authors suggest that also alternative sources of Se should be taken into account in the whole diet to counteract Hg’s potential effects [150]. Finally, the cooking methods of fish and seafood seem to be able to significantly reduce the Hg bioaccessible fraction, so that could be another factor to take into consideration for food safety regulations [151].

5. Conclusions

The results obtained in this work showed that the historical Hg mining in the Idrija district and the past CAP activity in the adjacent Friulian Plain are responsible for a huge and diffuse Hg contamination in the sediment of the MGL. The concentration of this element exceeds the threshold set for good quality sensu WFD/2000/60/CE and also those proposed for potential eco-toxicological effects. Overall, the MGL actually represents one of the most contaminated areas of the Mediterranean Sea, and remediation activities are mostly unfeasible due to the thickness of sediment involved in contamination and the delicate equilibrium of the system. In this context, the Hg speciation approach emphasises the limited mobility of Hg from sediment to the upper water column, resulting in a limitation in bioavailability for MeHg production. The prevalent species are mostly insoluble and, as found in the adjacent Gulf of Trieste, present a tightly bound mineral fraction consistent with cinnabar. The speciation of Hg explains the low values found on water surfaces, where the standard set by the WFD/2000/60/CE was never exceeded. It is clear that the hydrology of the MGL, characterised by low water residence times, justifies the absence of significant differences in concentrations in different areas of the lagoon environment.

The contamination of Hg in the sediment poses an important issue for the bioaccumulation of Hg in fish considering that the lagoon exploits fisheries and aquaculture. this work added to our understanding of a variety of species that are native to the area and are currently being collected for commercialization and consumption by local residents. Generally, the limit for commercialization is often overcome, especially in the Grado sector, which is the most contaminated part of the lagoon, and in the more confined areas subjected to more intense methylation processes, but this cannot be generalised because bioaccumulation is specie-specific and depends also on breeding season and sex. Notably, selected species with a detritivore feeding strategy display a significant correlation between total Hg content in sediment and edible parts. On the other hand, the calculated biota sediment accumulation factor (BSAF) suggests that the Hg bioavailable for the transfer from sediment to biota is little, and this justifies the fact that, despite the diffused Hg contamination, some areas are almost potentially safe for bioaccumulation. The consumption of fish, which is recommended for its high nutritional value, was also evaluated in detritivore species through the target hazard quotient (THQ). It appears that some parts of the lagoon seem to be safer for fish consumption.

Further investigations are needed to elucidate important steps of Hg bioaccumulation in organisms belonging to the lowest levels of the trophic chain (i.e., phytoplankton, epiphytes, seagrass, macroalgae, herbivorous benthos, and carnivorous macrobenthos), and regarding fish, more efforts should be conducted to investigate the content of Se due to its recognised counterbalancing effect.