Dataset on the major and trace elements contents and contamination in the sediments of Saronikos Gulf and Elefsis Bay, Greece

Coastal marine sediments receive intensive stress from urbanization and industrialization, which is manifested by increased contents of heavy metals and organic pollutants. Saronikos Gulf and the small embayment of Elefsis, stretch along the coast of the greater Athens and Pireaus port, the most urbanized and industrialized areas in Greece. Here we present the data of a 20-year geochemical record on grain-size, organic carbon, and major and trace elements contents of the Saronikos Gulf sediments. A total of 216 sediment samples were collected within the period of 1999–2018 from the four sub-sectors of the gulf, namely, the Elefsis Bay, the Inner, Outer, and Western (Megara and Epidavros basin) Saronikos Gulf. Additionally, at least one core was obtained from each sub-sector. Sediments deposited at pre-industrial periods were recognized by 14C and 210Pb dating, and served for establishing regionalized background levels of metals. Factor analysis was conducted to reveal the inter-parametric relationships, thus their common sources, as well as transport and deposition pathways. Then, Enrichment Factors and the multi-elemental Modified Pollution Index (MPI) were calculated to assess the current environmental status of the sediments. Data of sampling sites with at least a five-year record, were assessed for temporal trends, to explore whether sustained, increasing or decreasing trends of the MPI are observed. The dataset and analyses presented here support the research article entitled Geochemistry of major and trace elements in surface sediments of the Saronikos Gulf (Greece): assessment of contamination between 1999 and 2018 [1].

MPI are observed. The dataset and analyses presented here support the research article entitled Geochemistry of major and trace elements in surface sediments of the Saronikos Gulf (Greece): assessment of contamination between 1999 and 2018 [1].
© 2020 The Authors. Published by Elsevier Inc. This is an open access article under the CC BY license (http://creativecommons. org/licenses/by/4.0/).
Specifications Table   Subject Oceanography Specific subject area Geochemistry of sedimentary major and trace elements and pollution assessment Type of data Table  Chart Graph Figure  How data were acquired Sampling: Box corer, gravity corer; Major and trace elements: X-ray Fluorescence (PANalytical, PW-2400 wavelength dispersive spectrometer); grain-size: Sedigraph (Micromeritics 5100E, Micromeritics III PLUS); OC: CHN analyzer (Fisons type EA-1108); 210 Pb (Ortec EGamp; G)); 14 C AMS dating (Beta Analytics Inc.); data statistics (XLSTAT); Trend assessment: MAKESENS software Data format Raw Parameters for data collection Surface sediment samples (n ¼ 216) were collected over an extended sampling network in order to achieve spatial and temporal coverage of sedimentary geochemical data in the Eastern Mediterranean with respect to major and trace elements, together with grain-size and organic carbon parameters. Sediment cores were obtained from all sub-sectors of the study area to assess the regionalized background levels of metals and facilitate future contamination assessments in the Eastern Mediterranean's regional seas. Statistical analysis was conducted to assess the sources, and transport pathways of the variables studied. Temporal trend analysis was performed to reveal increasing or decreasing trend over the 20-years period of research. Description of data collection Sediment samples were obtained by means of a stainless steel box and gravity corer operating on board of the R/V Aegaeo. The samples were analyzed for grain-size, organic carbon and major and trace elements contents with the same analytical techniques (sieving and X-ray absorption, CHN analyzer, XRF, respectively). Statistical analysis, including the Kolmogorov-Smirnov normality test and Factor analysis on Box-Cox and z-score transformed data was carried out using Principal Factor Analysis with Varimax rotation (XLSTAT). Enrichment factors were calculated using as a background the local, pre-industrial levels, previously established by using radio-chronology data. A multielemental pollution index was selected after reviewing the recent literature and employed to the most recently obtained data set to assess the current environmental status. Data could be combined with other indicators (hazardous organic and inorganic substances in sediments, biota and water) to assess the overall quality status of the coastal and marine environment within the frameworks of WFD and MSFD. The dataset contributes to efforts of closing knowledge gap on metal levels in the Eastern Mediterranean Sea and could assist the development of the assessment criteria (background and background assessment concentrations) of Europe's regional seas.

Data description
Surface sediments (0e1 cm) were collected from 68 sampling sites in the Saronikos Gulf and its subbasins, namely, the Elefsis Bay, the Western (or Megara and Epidavros basin), the Inner and Outer Saronikos Gulf (Fig. 1a). Sampling was conducted by using a stainless steel box corer during thirteen oceanographic cruises of the R/V Aegaeo, from February 1999 to January 2018. A total of 216 samples were collected by using plastic tools and containers to avoid metal contamination. Several samples were obtained over the same network of stations serving the diachronic monitoring of the Saronikos Gulf with respect to major and trace elements contents. These samples were used to assess temporal trends of contaminants. Table S1 presents the grain-size, organic carbon, and major and trace elements contents of the surface sediments collected and analyzed from 1999 to 2018. The most recent dataset, i.e. the latest sediment samples (n ¼ 68) obtained throughout the sampling network of Fig. 1b, is given in Table S2. The latter dataset was studied in more detail to gain insight of the current environmental status of the study area. It consisted of samples collected from the Western Saronikos Gulf in 1999 (n ¼ 14); samples from the Elefsis Bay, the Inner and Outer Saronikos Gulf collected mostly from 2016 to 2018 (52 samples); two more samples were collected in the latter area in 2012e2013.
Furthermore, sediment cores were obtained from five locations. The K4a, S7 and SARC18 cores ( Fig.  1b) were retrieved using gravity corer, whereas, the S2 and S21 using a box corer. Sediment cores aimed at establishing the local background levels of major and trace elements. The recent sedimentation rates in two cores, S2 and K4a, were calculated using the 210 Pb method. The profiles of 210 Pb excess are shown in Fig. 2. The rest of the cores were dated by 14 C accelerator mass spectrometer (AMS) analyses. The profiles of metal contents normalized to Al for the five sediment cores, collected from the four subsectors of the Saronikos Gulf are shown in Fig. 3.
Chemometric analysis was conducted in the most recently obtained sediment samples (n ¼ 68) in order to explore inter-variable relationships and underlying common sources, as well as transport and deposition patterns. The variables of the dataset were first screened for normal distribution by the Kolmogorov-Smirnov (KS) normality test [2]. These results are given in Table 1. Transformed data to fulfil the requirements of normality [2,3], were introduced to Factor Analysis (FA). The analysis revealed three factors that explained 84% of the total variance ( Table 2; Fig. 4). The first factor (34.1%) showed positive loadings for clay, Si, Al, V, Mn, Co; F2 (30.6%) exhibited high positive loading for C org , Cu, Zn, As, and Pb; F3 (19.3%) associated Mg, Cr, Ni, Co, V, and Mn.
A two-step approach was followed for assessing pollution. First, Enrichment Factors (EFs) for the recent dataset of surface sediments were calculated according to Eq. (1): where background refers to the local pre-industrial levels established for each sub-sector of the study area by taking into account the results of 14 C dating and the recent sedimentation rates reported in Ref. [1]. The calculated EFs are presented in Table 3. As a second step for assessing contamination, the multi-elemental Modified Pollution Index (MPI), introduced by Brady et al. [4] was used. The calculated MPI values are given in Table 3.  Trend analysis was conducted to identify significant and sustained, increasing or decreasing trends of the MPI. Trend analysis was performed using the sampling sites with available data for more than 5 years, totaling 14 sites. Table 4 presents the output of the model.

Grain-size
Samples were wet sieved through a 63 mm stainless steel mesh. The finer fraction was analyzed with a Micromeritics Sedigraph 5100E, whereas samples collected after March 2017 with a Micromeritics III PLUS, in order to separate silt and clay fractions. Calgon (5.5 g L À1 ) and sonication (60 s) was used for disaggregation/dispersion. Subsequently, the dry percentages of sand (>63 mm), silt (2 mm <Ø <63 mm), and clay (<2 mm) were calculated and the nomenclature followed Folk [5].  Values in bold correspond for each variable to the factor for which the squared cosine is the largest.

Organic carbon content
Sediment samples were thoroughly ground in an agate mortar and very well homogenized to reduce variability between replicates. Splits of 10e20 mg of powdered homogenized sample were weighed accurately (0.01 mg) into specially designed silver containers. Organic carbon was determined after removal of inorganic carbon by acidification of samples with 20 mL of 6 N HCl at 60 C (this treatment was conducted five times in 12 hours intervals). After the inorganic carbon removal, the samples were dried at 60 C overnight. Then, the containers were pinched closed, compacted, and formed into a ball. The balls then were placed in the auto-sampler of a Fisons Instruments CHN elemental analyzer type EA-1108 to determine organic carbon contents. The operating parameters were very similar to those reported in Refs.
[6e8]. The precision of the method was within 5%.

Major and trace elements
Samples were sieved through a 1 mm sieve, oven dried at 40 C to remove moisture, and then ground to a fine powder in a motorized mill with agate mortar and balls. Sediment samples were analyzed for their chemical composition in a PANalytical (former Philips) PW-2400 wavelength X-Ray Fluorescence analyzer, equipped with Rh-tube. Major elements were determined in fused beads (SiO 2 , Al 2 O 3 , TiO 2 , Fe 2 O 3 , K 2 O, Na 2 O, CaO, MgO, P 2 O 5 ); for the purposes of the present paper we present only Si, Al, Mg and Ca contents and corresponding spatial distributions. Fused bead preparation involved a complete fusion of 0.6 g of sample, with 5.4 g of flux (50:50 lithium meta-borate, lithium tetra-borate) and 0.5 g of lithium nitrate, the latter being used as an oxidizer. Loss on ignition (LOI) was determined after burning 1 g of sample for 1 h at 1000 C.
Trace elements were determined according to the following procedure: 5 g of powdered sample were mixed with 1.25 g of wax and subsequently pressed in a 31 mm aluminum cup (20 s, 20 tons). The powder pellets were analyzed in the XRF to determine trace element contents (V, Cr, Mn, Co, Ni, Cu, Zn, As, Pb) using PANalyticals's Pro-Trace reference sample set and software for the instrument calibration. Analyses was based on 2-point calibrations forced through the origin, using 25 multi-element high quality standards and two blank samples, which allow the XRF system calibration for elements ranging from Sc to U; background, line overlap, and matrix corrections were applied. Analytical accuracy was checked by parallel analysis of the certified sediment standard PACS-2 and was found to be better than 7% for all elements analyzed. Analytical precision was checked in sample replicates and was always  better than 0.5%. Detection limits were below 5 mg kg À1 (V, Co, Ni, As), 5e10 mg kg À1 (Cr, Cu, Pb), and 10e12 mg kg À1 (Mn, Zn) for the elements determined (see also Table S1). To check long-term repeatability in the framework of the present study, archived powder samples were re-scanned.

Recent sedimentation rates
For the calculation of the recent sedimentation rates, the down core total 210 Pb activity was determined through the activity of its alpha-emitting granddaughter 210 Po, assuming secular equilibrium with 210 Pb. For the total dissolution of the dried sediments the analytical method described by Sanchez-Cabeza et al. [9] was applied. The sedimentation rates were calculated using the widely used Constant Rate of Supply model (CRS) [10].

Factor analysis
Factor analysis is a powerful method in geochemistry that explains the variation in a multivariate data set by a limited number of factors [2], provided that certain precautions are taken. The data set (most recent data set; n ¼ 68) was carefully checked for outliers by means of Box-and-Whisker plots, and a few stations were removed from FA analysis due to the presence of several univariate outliers. Since FA is based on the correlation (or covariance) matrix, it is sensitive to non-normally distributed variables, thus various transformations are required prior to analysis [2]. Normal distribution of all variables was tested by a Kolmogorov-Smirnov (KS) normality test. Eight out of 18 variables did not pass the KS test run on the original data, whereas four variables did not pass the KS test run on lognormalized data ( Table 2). A Box-Cox transformation followed by a z-transformation [3] provided the best results, where all variables but sand, and silt passed the KS test; grain-size variations were therefore explained by the clay fraction variability. Subsequently, Principal Factor Analysis with Varimax rotation was conducted in order to group clay, organic carbon and major and trace elements.

Assessment of temporal trends
The MAKESENS program has been developed by the Finnish Meteorological Institute [11] for detecting trends of annual values of atmospheric pollutants. MAKESENS performs two types of statistical analyses: First, the presence of a monotonic, increasing or decreasing trend is tested with the nonparametric Mann-Kendall test, and second, the slope of a linear trend is estimated with the nonparametric Sen's method [12].
The Mann-Kendall test is applicable in cases when the data values x i of a time series can be assumed to obey the model: where: f(t) is a continuous monotonic increasing or decreasing function of time, and ε i are the residuals that can be assumed to be from the same distribution with zero mean.
The null hypothesis H 0 is that observations x i display no trend and is tested against the alternative hypothesis H 1 that there is an increasing or decreasing monotonic trend. For time series with less than 10 observations (as in this study), the S statistics is calculated by the formula [12]: where: n is the number of annual values in the studied data series, x j and x k are the annual values in years j and k, respectively, and j > k, and For n 9 the absolute value of S is compared to the theoretical distribution of S derived by Mann and Kendall [12]. In MAKESENS the two-tailed test is used for four different significance levels a: 0.1, 0.05, 0.01 and 0.001. At certain probability level H 0 is rejected in favour of H 1 if the absolute value of S equals or exceeds a specified value S a/2 , where S a/2 is the smallest S which has the probability less than a/2 to appear in case of no trend. A positive value of S indicates an upward trend, and a negative value indicates a downward trend [11]. The non-parametric Sen's method is used to estimate the true slope of an existing trend. It can used in cases where the trend can be assumed to be linear. Thus, the f(t) in Eq. (3) could be written as: where Q is the slope and B is the constant.
To calculate the slope estimate Q of Eq. (6), the slopes of pairs of data values are calculated first: where j > k.
For n values of x j and N data pairs for which j > k, the Sen's estimator of slope will be the median of these N values of Q.
The constant B of Eq. (6) is the median of the n values of differences x i e Q ti .