Heavy Metalspatial Distribution and Pollution Assessment in the Surface Sediments of the North – Western Black Sea Shelf

22 surface sediment samples were collected in August 2018 from the Romanian inner shelf (Nord-Western Black Sea). Concentrations of some metals (Al, Cr, Cu, Ni, Zn, As, Pb, and Hg), TOC content, and grain size of sediment samples were determined by specific techniques. The order of accumulation of heavy metals was Zn>Cr>Ni>Cu>Pb>As>Hg. Multivariate analysis indicated that As, Ni, Cu, Zn, Pb, and Hg concentrations had similar behavior and they were positively correlated with the clay content, whereas Al and Cr concentrations presented close patterns and they were negatively correlated with the water depth. Sediment pollution assessment indices (enrichment factor, contamination factor, and geo-accumulation index) suggested no/low pollution for most of the metals analyzed, excepting for Pb and Hg (moderate pollution). Values of pollution indices highlighted a higher sediment pollution with Pb and Hg along the Danube’s plume direction, in the oil platform area (eastern edge of the Portita Bay), and partially in the Constanta and Mangalia area, suggesting the influence of port activities, tourism, urban wastewater discharges, oil and gas extraction.


Study area
Sediment samples were collected from 22 sampling stations (water depths within 12-67 m), covering the Romanian inner shelf waters (NW Black Sea), during the research cruise aboard R/V Mare Nigrum conducted in August 2018. The stations considered in this study were near Sulina (SU01 and SU02), Sfantu Gheorghe (SG01, SG03-05), Portita (PO01, PO02, PO04, and PO05), Constanta (CT01-05), Eforie (EF02), Tuzla (TZ18), and Mangalia (MA04-08). Spatial distribution of selected stations is considered representative for the assessment of the metal pollution along the Romanian coast of the Black Sea. The map of the sampling stations considered in this study is shown in Figure 1, while their main characteristics in terms of coordinates and water depths are summarized in Table 1.

Sediments sampling and analysis
Sediment samples were collected from the surface layer (0-2 cm depth) using a grab sampler with an opening mouth of 0.14 m 2 . Each sample was subsampled for grain size and geochemical analyses, respectively. Sediment subsamples were stored in plastic bags and kept at 0-4 °C until subsequent analyses.
Grain sizes of sediments were determined with a Mastersizer 2000 laser diffraction granulometer (Mastersizer 2000E, ver. 5.20) and associated dispersion units (Malvern Instruments, U.K), measurement precision being of 1% and result reproducibility below 1%. Separate granulometric fractions are in accordance with Udden-Wentworth dimensional scale with sand/silt and silt/clay boundaries of 63 µm and 4 µm, respectively. The Shepard's ternary diagram was used for lithological classification of sediment samples [21].
Before geochemical analyses, the sediments were oven dried (24-48 h/105°C), ground, and homogenized with a mortar and pestle. TOC concentrations were determined using titration method [22].
Concentrations of Al, Cr, Cu, Ni, Zn, As, and Pb were measured by X-ray fluorescence spectrometry using an EDXRF Spectro Xepos spectrometer (Germany). Total Hg content was determined using an automatic mercury analyzer DMA 80 Milestone (Italy) by solid sample thermal decomposition, identification, and quantification of total Hg by atomic absorption spectrophotometry. To validate the analytical methodology, a certified reference material NCS DC 73022 was used. Measured and certified values of element/compound concentrations were compared (Table 2). For this material, all measured values were statistically similar to the certified values (p<0.05), demonstrating the reliability of the methodology and the estimated concentrations.
Pollution indices were determined based on Eqs.  Table 3.

Experimental data
Composition (expressed as percentages (P) of sand, silt, and clay) and type of surface (0-2 cm) sediments in the area considered in the study are specified in Table 4. Depending on their composition, the sediment types vary from silty sand to clayey silt.

Table 4. Composition and type of surface sediments
The values of TOC and metal concentrations in the surface sediments for each station are specified in Table 5. Descriptive statistics in terms of minimum (min) and maximum (max) values, mean, median, standard deviation (SD), and coefficient of variation (CV) for TOC and each metal concentration are also summarized in Table 5.
Data presented in Table 5 and Fig. 2 highlight TOC concentrations in the surface sediments of the investigated area between 0.092% and 1.779%, with a minimum in front of the Constanta harbor (CT01) and a maximum in the deepest station from the Portita Bay (PO04). Lower TOC concentrations (<0.5%) were observed at stations PO01 and MA05, while quite high concentrations  56.9% (for Hg). Spatial distributions of metals in the surface sediments, which are shown in Fig. 3, highlight the following issues: (i) the highest values of Al concentration, i.e., 60640-64930 mg/kg, were detected in the Danube's mouth area (SU01, SG01, and SG03), due to the strong influence of the Danube's input, whereas minimum values (19910 and 22510 mg/kg) were found in the deepest stations from the southern part of the studied area (MA04 and CT05); (ii) the highest values of Cr concentration (>90 mg/kg) were noticed not only in the Danube's mouth area (SU01, SG01, and SG03), but also in the Portita Bay (PO02, PO04, and PO05) and  Previous works carried out in the period 1997-2007 showed also a large spatial variability of the heavy metals in the surface sediments of the Romanian shelf [25]. Generally, the metal concentrations determined within 1997-2007 were quite higher than the values measured in this study, e.g., cCr,s=34. 17

Statistical processing of experimental data
PCA, which was performed on 8 metal (Al, Cr, Ni, Cu, Zn, As, Pb, and Hg) concentrations, TOC content, water depth, percentages of sand, silt, and clay, highlighted two principal components (PCs) with eigenvalues greater than 1, which account for 84.51% of total variance (TV). The results shown in Figure 4 and Table 6 emphasize the following aspects: (i) PC1, explaining 61.96% of TV, is characterized by high negative loads for As, Ni, Cu, Zn, Pb, Hg, and clay contents as well as a high positive load for sand percentage; (ii) PC2, accounting for 22.56% of TV, presents a high positive load for water depth (h) as well as high negative loads for Al and Cr concentrations. Pearson correlation coefficient matrix (Table 7)  positive correlation between Al and Cr contents (cAl,s and cCr,s) as well as high negative correlations between h and cAl,s and h and cCr,s. Accordingly, As, Ni, Cu, Zn, Pb, and Hg concentrations have similar behavior and they are positively correlated with clay content, whereas Al and Cr contents present close patterns and they are negatively correlated with water depth. More studies in the related literature pointed out that clay particles are important carriers of heavy metals as well as that the water depth can have a significant effect on spatial distributions of metals in the surface sediments [4][5][6][7].
Data depicted in Figs      Data shown in Figure 5 indicate good discriminations between clusters I and III on the PC1 direction and between clusters I and II on the PC2 direction.

Pollution indices
Descriptive statistics (minimum, maximum, mean, median, standard deviation, and coefficient of variation) related to pollution indices corresponding to each metal, i.e., EF, CF, and Igeo, which were estimated using Eqs. (1)-(3), are summarized in Table 8.
. Spatial distributions of EFi (Figure 7) show values between 0.27 and 4.39 (no, minor, and moderate pollution), with a minimum for Cu at the station EF02 and a maximum for Hg at the station SG05. ANOVA one way test followed by Tuckey (HSD) multiple comparison test revealed significant higher values only for Pb and Hg.
Values of EFi greater than 3, suggesting moderate enrichment, were found only for Pb (maximum of 4.36) and Hg (maximum of 4.39). Higher values of EFPb and EFHg found at stations SG04 (3.66 and 3.61), SG05 (3.03 and 4.39), CT04 (2.97 and 3.34), and CT05 (4.36 and 4.00), which are situated on the Danube's plume direction, can be related to the riverine discharges. Moreover, higher levels of EFHg (3.42-4.03) were detected at stations located in the Portita Bay (except PO01), most probable linked to the oil and gas extraction activities in this area, at station SG01 (3.31), due to the direct influence of the Sf. Gheorghe branch, and also at station MA08 (4.03), possibly related to the local regime of currents carrying on the wastewater and industrial discharges.
Lower levels of EFPb (1.13-1.30) and EFHg (0.43-0.79) were observed at shallow stations PO01, CT01, EF02, and MA06, characterized by lower clay contents (4.50-11.43%) and higher sand percentages (36.75-66.02%). Moreover, the lowest values of EFi for Pb (1.13) and Hg (0.43) were found at the station EF02, from which sediments with the lowest clay content (4.50%) and the highest sand content (66.02%) were collected. These findings suggest the influence of the sediment type on the pollution level. All

Conclusions
Spatial distributions of some trace elements (As, Cr, Ni, Cu, Zn, Pb, and Hg) and Al in the surface sediments collected from NW Black Sea were determined. Mean values of heavy metal concentrations (0.11-74.11 mg/kg) decreased in the order: Zn>Cr>Ni>Cu>Pb>As>Hg. The elements considered in this study showed relatively large spatial variability with higher concentrations in front of the Danube's mouths and at eastern edge of the Portita Bay.
Multivariate analysis revealed that As, Ni, Cu, Zn, Pb, and Hg concentrations had similar behavior and they were positively correlated with the clay content, whereas Al and Cr contents presented close patterns and they were negatively correlated with the water depth. Three clusters of stations having dissimilar metal accumulation in the surface sediments were obtained.
Different indices used to evaluate the degree of pollution of the Romanian inner shelf sediments suggest that most metal concentrations in the sediments were, generally, at natural levels. Dominant heavy metal pollution in the Romanian inner shelf sediments came from Pb and Hg. The values of EF, CF, and Igeo indices suggest a higher sediment pollution with Pb and Hg along the Danube's plume