Spatial contamination and health risks of heavy metal(loid)s in surface soils from a petrochemical complex in the north-eastern region of Algeria

The spatial distribution patterns, origins and environmental and health risks of metal trace elements (As, Cd, Co, Cr, Cu, Ni, Pb, Zn) have been evaluated through spiking and analysis of 84 soil samples in the region of Arzew (Algeria). This city has one of the biggest petroleum harbours of Algeria and in the same platform several industries, coexisting with residential and agricultural areas. The mean concentration of all the metals studied exceeds their corresponding background levels and/or exceeds the ecological predicted no effect concentration values. The geo-accumulation and the potential ecological risk indices indicate a low to moderate risk for As, Co, Cu, Ni and Zn and a moderate to considerable risk for Cr and Pb. According to the health risk assessment results, the carcinogenic risks due to Cr exceed the acceptable level in all the functional areas, and when considering combined effects of metals, the hazard index obtained shows a high potential carcinogenic risk to the local children.


Introduction
Soil metal contamination is of great concern for human and environmental health, particularly in areas submitted to a concomitant increase in industrial activity and a growing population. This is particularly valid in places where expansion of industry has been chaotic resulting from an absence of planning, and where, today, dangerous undertakings are not separated from residential areas, school zones, agricultural areas, or either leisure activities. In such areas, Metal Trace Elements (MTE) are released partly from vehicular emissions (due to wear and tear of vehicle parts such as brake, tire and clutch), and partly from agricultural activities from excess use of amendments (e.g., sewage sludge and wastewater, animal manure, composts) or agrochemicals (e.g. fertilizers) (Solgi et al. 2018). Another important part of metal source in soils, after geochemical or natural background, comes from industrial activities, issued from smelting, mining, metal processing, dyes, pigments, ceramic, tanning, textile industries (Abraham et al. 2018), and from petrochemical industries.
Petrochemical industries have been identified as important emission sources of inorganic and organic pollutants that can pollute soils, both in close proximity of petroleum refining and petrochemical manufacturing plants (by improper disposal or bad treatment of effluents), and at distal areas through the waterborne transport of fine particulate metals, and together with windborne dispersal of dry fine particles, with a diffuse contamination gradient decreasing with distance from the source (Karim et al. 2015).
Elements such as cadmium (Cd), chromium (Cr), copper (Cu), nickel (Ni), lead (Pb), vanadium (V), zinc (Zn), cobalt (Co), manganese (Mn), molybdenum (Mo) and arsenic (As) among others, can be released to air and can thus be inhaled by workers but also by nonoccupationally exposed populations living in the neighbourhood. Once deposited, these elements can bioaccumulate in plants and animals and may eventually reach humans through the food chain via direct intake of heavy metals by plants from soils (Bermudez et al., 2011).
Drinking waters are another mode of ingestion of MTE, especially in heavily polluted areas . Direct oral soil ingestion cannot be ignored: soil pica behaviour has been implicated in several case studies in which MTE exposure in children with elevated MTE levels in soils were correlated. After ingestion and inhalation, skin contact constitutes the third main way of human exposure to soil metals (Cao et al. 2018).
Even some metals such as Co, Cr (III), Cu, Mn, Mo and Zn are considered as essential compounds for humans at low concentrations due to their roles in metabolic functions (Prashanth et al. 2015), and even Ni and V are supposed to have beneficial biological functions in plants and some animals, the continued exposure to low levels of MTE may lead to adverse health effects. According to the International Agency for Research on Cancer (IARC), As, Cd, Cr (VI) compounds and Ni are classified as carcinogenic, whereas Co, Pb and Sb trioxide are considered as possibly carcinogenic (Mulware, 2013). The organs affected by these metals are kidney, lung, liver, gastrointestinal and haematological systems, mainly the peripheral and central nervous systems (Karri et al. 2018) though their entry into the food chain or by the inhalation of dust from contaminated surface soil (Tang et al. 2017).
Moreover, exposure to MTE may produce additive or synergistic interactions or even new effects that are not seen in single component exposures (Kossowska et al. 2013).
Some epidemiological studies have also pointed out the possibility that living adjacent to petrochemical facilities could induce an excess of leukemia, as well as bone, brain, and bladder cancers and preterm deliveries (Barazza et al. 2018). However, the detailed pathogenic mechanisms remain unclear due to the complex mixture of pollutants emitted by the petrochemical industry, which include not only MTE but also sulfur oxides, nitrogen oxides, metals, polycyclic aromatic hydrocarbons, and volatile organic compounds (Kampeerawipakorn et al. 2017).
The region of Arzew is located between Stidia and Mostaganem towns in the Algerian shoreline of Mediterranean Sea (Figure 1), about 40 km east from the city of Oran and at 400 km West from Algiers. It constitutes one of the biggest petroleum harbours and many industrial complexes (gathering oil, gas, refining and petrochemical industries) are situated in this area making it an industrial and commercial hub for Algerian economy. The first large-scale Liquefaction Natural Gas (LNG) plant in the world started began functioning in 1964 when LNG was transported from Arzew, Algeria to Canvey Island, with initial production capacity of 2560 metric tons a day of LNG (Tayeb et al. 2015). The industrial platform is spread over an area of 2500 ha. The pole consists of a large number of hydrocarbon and petrochemical plants, liquefaction natural gas, liquefied petroleum gas, a refinery, three units of ammonia and nitrogen fertilizers production, a methanol synthesis complex, an electricity generation plant, and a seawater desalination plant.
There are very limited published data on the soil contamination in Arzew, existing published studies being focused on the impacts of petrochemical activities on marine ecosystems (Dauvin et al. 2017), whereas a large part of MTE found in Gulf or Arzew come from inland (from leaching of soils, and by wet or dry atmospheric deposition). From an environmental and human health point of view (the petrochemical platform is surrounded by urban concentration of about 148,782 inhabitants), it seems thus important to monitor the impact of anthropogenic activities on heavy metal contamination of urban soil and for the development of proper management strategies for urban environment pollution control and for the remediation of heavy metal contaminated soils of Arzew city.
The main objectives of this study were thus to estimate the MTE concentrations in soils of Arzew and describing their distribution patterns according areas defined by human activities by use of pollution indices (geo-accumulation index and potential ecological risk index).
Furthermore, it aimed at assessing the potential health risks associated with MTE for children living in this city, by calculation of hazard quotients for inhalation, dermal and ingestion exposures, and by the determination of resulting hazard index. This study intends thus to provide for the first-time results on the multi-contamination of soils of Arzew, where industrial and heavy road traffic activities coexist with agriculture and residential areas and intends to be useful for local authorities for the development of risk management measures.

Study Area
The study site covers the Arzew petrochemical platform area, that covers a part of the two   Arab et al., 2015); 2b. Pedological map issued from the general geology map designed by Fisher et Doumergue (1900) and locations of the sampling stations (S1 to S6) The soils are mainly constituted of clays and sand-rich alluvium issue from run-off deposits, and of sandstone and limestone ( Figure 2b; Ficheur et Doumergue 1908). Two main zones can be distinguished pedologically in the study sites: quartz-sand and humus-bearing soils, at the level of the dune coastal and sublittoral area, and alluvial soils with predominance of clay texture dominates. The climate is typical of a Meditteranean area, with mild temperatures throughout the year (data observed these last ten years: minimum: 15.4 °C; maximum : 23.6 °C; mean : 18.5 °C). Precipitations occur mainly between October and February, with average annual rainfall around 324 mm.

Sample collection
A sum of 84 topsoil samples (0-20 cm) were collected from six functional sections around the petrochemical platform of Arzew, representing a triangular area 17 km wide by 12 km high.
These six sections have been chosen to represent the main activities carried out by people working and/or living in this area. 9 samples were collected from coastal area (S1, beach zone), 13 collected from port area (S2, harbour zone with private recreational vessels and were also collected from an area 15 km away from the petrochemical platform area (presumably unpolluted zones) and representing the two pedogeochimical soils of the study area, for background studies. All the sampling points were systematically distributed in the study area, recorded with a GPS and, collected in clean polythene bags to avoid contamination before analysis in the laboratory.

Sample preparation and analysis
Soil samples were sieved to 2 mm mesh, air-dried at room temperature and then ground (RETSCH zm 1000 with tungsten blades) to pass through a 0.2 mm titanium sieve before analyses. The acid digestion of the sampled soils (0.5 g) was carried out in a microwave mineralizer (Milestone Start D Microwave Labstation, Sorisole, Italy) using aqua regia (1/3 HNO3 + 2/3 HCl). Suprapur grade 65 % (m/m) HNO3 and 37 % (m/m) HCl (Merck, Germany) were used for sample mineralization. This microwave digestion system was equipped with ten 100-mL tetrafluoromethoxy vessels and a ceramic vessel jacket. To protect the unit, the cavity and the door were plasmacoated with Polytetrafluoroethylene (PTFE).
After digestion and cooling, the mineralized products were filtered with a 0.45-μm mesh and diluted to a final volume of 25 mL with ultrapure water. The metal contents were determined by Inductively coupled Plasma-Atomic Emission Spectroscopy (ICP-AES, Jobin Yvon Horiba, Spectra 2000) for As, Cd, Co, Cr, Cu, Ni, Pb, and Zn. Monoelemental high-purity grade 1 g L -1 stock solutions of As, Cd, Co, Cr, Cu, Ni, Pb, and Zn were purchased from Merck (Darmstadt, Germany). The purity of the plasma torch argon was greater than 99.99 %.
All laboratory glassware was soaked in a 10 % (v/v) HNO3 solution bath for 24 h and was rinsed with high-purity water, then dried under clean air conditions at ambient temperature.
All reagents were of the highest commercially available purity grade. Deionized water (resistivity 18 MΩ cm -1 ) that was obtained from a Milli-Q purification system (Millipore, Molsheim, France) was used to prepare all standard and sample solutions.

Statistical analysis
Relationships between MTE and their potential emission sources have been assessed by performing Pearson's correlation analysis and principal component analysis (PCA), by use of the commercially-available statistics software IBM SPSS version 24.0.0 for windows.

Pollution indices
The accumulation levels of individual MTE in soil samples were evaluated by use of the geoaccumulation index (Igeo) and were calculated as presented in the Annex 1 in the supplementary material.
The potential ecological risk index (PERI) associated with MTE in soil samples has been calculated according Håkanson's methodology (1980). The geoaccumulation index (Igeo) was calculated as follows: where Cn is the concentration of the metal measured in soil and Bn is the corresponding geochemical background value. These latter have been chosen in BG1 and BG2 according texture of soil considered. The constant 1.5 is introduced in this equation to minimize the variations due to lithogenic actions and effect of possible variations in the background values (Pathak et al, 2015). The Igeo for each metal is calculated and typically classified as follows: unpolluted (Igeo≤0), unpolluted to moderately polluted (0 < Igeo≤1), moderately polluted (1 < Igeo≤2), moderately to heavily polluted (2 < Igeo≤3), heavily polluted (3 < Igeo≤4), heavily to extremely polluted (4 < Igeo≤5), and extremely polluted (Igeo > 5).
The PERI of each MTE was defined as being: where Tr i is the toxic response factor of each metal (As = 10, Ni = 6, Pb = Cu = Co = 5, Cr = 2, Zn = 1), C0 i is the concentration of each metal, and Cn i is the corresponding background value. The Er i is classically classified following: low ecological risk (Er i ≤ 40), moderate ecological risk (40 <Er i ≤80), considerable ecological risk (80 <Er i ≤160), high ecological risk (160 < Er i ≤320), and very high ecological risk (Er i >320).

Non-carcinogenic risk assessment
In this study, only children have been considered for the assessment of the probability of noncarcinogenic risks due to the presence of MTE in topsoils of Arzew. This public is exposed to MTE through the three main pathways that are direct ingestion (pica-behaviour not considered), inhalation of soil particles through nose and mouth, and dermal contact absorption. The human non-cancer risk effects from metals can be assessed using the hazard quotient (HQ), which is the ratio of the average daily dose (ADD) to the reference dose (RfD, mg kg -1 day -1 ) of a MTE for the same exposure pathway. These RfD used for the calculation of HQ are given in Table 1; they were taken from U.S. EPA Exposure Factors Handbook (2011), from the Risk Assessment Information System (RAIS) website. The ADD is a dose averaged over a specified period. The general equation used for ADD is (U.S. EPA, 2014): Where: ADDsoil ing is the potential subchronic average daily dose of the contaminant from ingestion of contaminated soil (mg kg -1 day -1 ); Csoil is the concentration of the contaminant in the ingested soil (mg kg -1 soil); CF is the conversion factor of 1.10 -6 kg/mg; IRsoil is the ingestion rate of soil (mg day -1 ). According to U.S. EPA recommendations (2011), the recommend central tendency ingestion rate of soil for children is 50 mg day -1 ; EF is the exposure frequency (day year -1 ). In this calculation, this exposure frequency is set up at 350 days, assuming that children are away from home (from the source of contamination) for two weeks per year; ET is the average time spent outside and playing at the contact of grounds. U.S. EPA (2011) estimates this duration at 202 min day -1 , by addition of the duration spent on playing on sand or gravel, on grass and on dirt. ET is thus equal to 0.14 (202 min/24h); ED is the exposure duration (year). Considering children as exposed populations, the duration has been set up at 18 years; BW is the average body weight (kg) on the considered period (18 years When ADD is used to determine a dermal exposure, additional terms must be included (surface area, adherence factor or dermal permeability coefficient) and similarly when considering inhalation exposure (emission factor) where EF is the exposure frequency (350 days per year), ED is the exposure duration (18 years), ABS is the absorption fraction (0.1, assuming that 10% of MTE in the soil contacting the skin is absorbed), SA/BW is the surface area of the skin that contacts the soil (only hands -0.057 m 2 -, arms -0.176 m 2 -and legs -0.358 m 2 -have been considered as being exposed to contaminated soils for the calculations) per kg of body weight (41 kg), AFsoil is the adherence factor for soil (mg cm -2 , depending on the body part considered; 0.11 0.011 and 0.031 mg cm -2 , respectively for hands, arms and legs) and AT is the averaging time (6300 days).
The sum of the HQ (HQingestion + HQdermal + HQinhalation) represent the hazard index (HI), that allow to estimate the potential non-carcinogenic effects following the value of HQ. If HI ≤1, the exposed public is unlikeky to experience adverse health effects, whereas opposite effects are expected if HI > 1 (U.S. EPA 1989).

MTE concentrations
Descriptive statistics of MTE concentrations (in mg kg -1 ) in topsoils from the six sectors in Arzew and surroundings are summarized in Table 2, together with soil background values and French regulatory limits (AFNOR 1996). levels of Cd were very low with values below 0.01 mg.kg -1 in all the samples and will not be discussed further. Globally, data obtained follow a normal distribution except for the road area (S6) where values of standardized skewness and standardized kurtosis are not within the range expected for data from a normal distribution (-2, +2). This zone is the longest one among areas studied and the traffic is quite different between road N11 (mostly cars) and road N13 (mostly trucks). For the other zones and whatever metals considered, the coefficients of variation are relatively low and indicate a homogeneous contamination in all soil samples.
Total As concentrations of 95% of the samples were under 20.0 mg kg -1 , the median value being 10 mg kg -1 . A recent study carried out in 33 countries in Europe (Reimann et al. 2018) indicate that more of 95% samples soils presented a geochemical threshold value below 20 mg kg -1 . In China, geochemical reference value relative to As is 8.40 mg kg -1 (CNEMC 1990). In our sampling points considered as being unpolluted (BG1 and BG2), arsenic is highly below these guideline levels (around 4 mg kg -1 ). However, results of independent sample test (p<0.05) showed that concentrations of As is significantly higher in the S1 and S2 areas (mean: 15.63 mg kg -1 ; max: 48.06 mg kg -1 ) and exceed the French regulatory limit (37 mg kg -1 ; AFNOR U44-041). Arsenic in soils may originated from not only natural sources  such as erosion or weathering of rocks, but also from anthropogenic activities including industrial emissions, application of pesticides, sewage irrigation, and atmospheric deposition . At the opposite, levels of As are comparatively relatively low in S5 (agricultural area), suggesting that As-rich fertilizers or pesticides have not been intensively used up to now in this area (Figure 3). Because of location of higher contaminated sampling points in S2, and because this area is under prevailing southerly winds, As is supposed to come from the industrial area (S5). Concentrations of Co observed in the soils of Arzew and its surroundings ( Table 1) are in agreement with pedogeochemical background values. 100% of the samples are below the French regulatory limit concerning Co (30 mg kg -1 ). Mean Co levels are 10.13 mg kg -1 (min: 3.43; max: 21.83) and relatively close to the BG1 and BG2 levels ( Table 1), with at least an order of magnitude three higher than these background levels. The mean levels of Co were similar between S1, S2, S3 and S6 (around 12 mg.kg -1 ), and S3 and S4 (around 7 mg. kg -1 ).
Dissemination of cobalt in the environment is mainly caused by mining activities, smelting and industrial processing and incineration of combustible municipal solid waste (Leyssens et al. 2017). From these results, it can be suggested that the petrochemical platform does not emit cobalt in the atmosphere (lowest level of Co in S4), or a high dispersion due to the wind or a low retention of Co in the topsoil.
Total Cr concentrations have a median of 60 mg kg -1 with a high range (14.42 -193 mg kg -1 ).
Only one sample exceed the French regulation limit (150 mg kg -1 ) and is in the industrial area (S4 , Table 1 and Figure 3). However, the whole sampling sites (except the agricultural area) present a relatively high level of Cr, with 90% of samples having Cr at more than 80 mg kg -1 , representing levels eight times higher than BG2. These levels exceed also the (eco)toxicological PNEC values derived by the European REACH regulation ( Table 1).

Origin of Cr input into the soils may come from phosphate fertilizers and metallurgical
industries (Maas et al. 2010). The diffuse Cr concentrations measured in the soils of our sampling sites suggest that agricultural activities may use Cr-containing materials (but statistically this area is not contaminated by Cr), and more probably that chromium found in the whole sampling sites may be originated from the industrial platform of Arzew. However, presence of Cr due to the road traffic cannot be excluded; indeed, its presence has already been associated with the bulk matrix from road dust, and yellow road paint and road pavement are typically sourcing of Cr (Thorpe and Harrison 2008).
Total Cu concentrations have a mean of 25.91 mg kg -1 and a median of 20.77 mg kg -1 .
Concentrations are globally homogeneous within the study area, although there are two functional areas that differ significantly, namely S5 in its southeastern portion and S6 in its western portion. In these latter portions, Cu levels are significantly higher than in the other areas, with levels higher than 60 mg kg -1 (max: 63.31 in S5; max: 223.7 in S6) ( Table 1 and Figure 3). In S5, the presence of copper is associated to the use as a post-harvest fungicides of Bordeaux mixture (made of lime and copper sulphate) on orange trees. In S6, the presence is associated to the vicinity of the road traffic. Cu is known to be abundant in brake lining materials, to be used in road pavement applications and could thus be found in particulate matter and dust in the urban environment (Men et al., 2018).
Ni presents the same pattern of geographical distribution as Co, but with higher levels ( Figure   S1). Total Ni concentrations have a median of 28.54 mg kg -1 and a mean of 26.94 mg kg -1 (min = 5.06; max = 51.84). 99% of the samples are below the French regulatory limit (50 mg kg -1 ) and 50% are below the PNEC value (65 mg kg -1 ). The highest concentrations are found in the coastal area (S1), in the harbour one (S2) and along the road zone (S6), with some sampling points having Ni at more than 50 mg kg -1 . Ni is widely used in various industrial sectors such as refining industry, textile and painting industry, foundries and metal processing industries (Aden et al. 2017).
Total Pb concentrations have a median of 52.11 mg kg -1 with a high range (13.27-220.5 mg kg -1 ) ( Table 1). Average concentrations are 63 mg kg -1 in the study area but the highest levels are found in S6, area representing the road zone, with 50% of samples (N=27) exceeding 100 mg kg -1 , which is the French regulatory limit for this MTE. Such an increase in Pb concentrations in areas with dense traffic has been observed in other cities, even with the sharp increase of unleaded fuel utilization in European countries because of the high stability of lead in soils (Turner and Lewis 2018). Moreover, in Algeria 89% of the gasoline consumption is still leaded (Maas et al. 2010). Therefore, the locations of Pb, together with absence of lead in the surroundings of the industrial platform, let us think that road traffic is the most likely sources of Pb in these soils.
None of the measured Zn concentrations exceed 300 mg kg -1 , which is the PNEC for value for this metal ( Table 2). Their median and mean concentrations are 83.34 and 93.27 mg kg -1 , respectively. The highest concentrations are found in the industrial area (S4) and along the road (S6), with 80% samples and 30% samples being at levels above 100 mg.kg -1 , respectively. Zn present in roadside soils is known to be issued from tire abrasion, lubricating oil, vehicular exhaust and breakpad wears (Morse et al. 2016) and atmospheric deposits of Zn to topsoils are mainly associated to municipal waste incinerators coal combustion residues, spillage of petrochemicals, mining and smelting industries and wildfires (Mousavi et al.

Correlation between metals
Statistical correlation analysis may provide an efficient method to understand the influencing factors and the sources of chemical contaminants. High correlations between MTE in soil may reflect that the accumulated concentrations of these metals come from similar emission sources. The results of the Pearson's correlation coefficient and their significance levels (p < 0.01) are presented in Table 3

Principal component analysis
PCA can be conducted to identify the potential sources of contamination (Tang et al. 2017), which may be responsible of the spatial distribution patterns of MTE. The Kaiser-Meyer-Olkin Index was calculated at 0.605 and results of the Bartlett's sphericity test were significant at p < 0.01, confirming that MTE concentrations were suitable for PCA. According to the results of the initial eigenvalues (higher than 1.00), four principal components are considered and explain 75.551% of the total variance (Table 4) The results of the matrix of the PCA together with its loading plot are shown in Figure 4.  Table   2), suggesting similarities in their sources. These three MTE are also mostly located in zones S1, S2 and S6 (in its North western portion) and their highest concentrations are found in S4 (industrial area), suggesting a main emission source from the industrial platform and a dissemination of metallic particles from South to North in the study area.
In the loading plot (Figure 4) However, we observe that the highest mean is found in the industrial area (S4), whereas the highest concentration is present in the harbour zone (S2), suggesting -as for Co, Cr and Ni-an emission source from the industrial platform towards the harbour zone.
Cu and Pb are linked together in PC4 (14.8%) with respective loadings value of 0.644 and 0.633. As seen before, these two metals are also distributed in PC2 (Pb) and in PC3 (Cu). This suggest that these metals are controlled by more than one factor and they originate from mixed sources, such as industrial activities and road traffic. Figure 6 presents the box plots of the Igeo in soil samples as a function of the functional areas under study. In all the areas studied, the mean Igeo are found between 0 and 1.00 for As, Co, Cu and Zn, indicating a low pollution level due to these MTE. The average Igeo values for Cr, Ni and Pb is found between 1.00 and 1.50, indicating a moderate pollution with these MTE. The ends of the whisker are set at 1.5 × IQR (interquartile range) above the third quartile and 1.5 × IQR below the first quartile. When the maximum values are outside this range, then they are shown as outliers Figure 6 : Box plots of the Igeo for As, Cr,Cu, Co, Pb, and Zn in soil samples as a function of the functional areas under study Also noteworthy is the fact that Igeo value for Cr exceed 1.0 in 71.6% of the whole samples, and even 2.00 in 45% of the samples, with 63% exceeding 2.00 in samples issued from S6, 46% from S4, and 37.5% from S1. The same observation can be made with respect to Pb with the finding of the 1.00 value is exceeded in 100% of the samples issued from S6, and where Igeo are beyond 3.00 in 14,8% of the samples.

Pollution indices
In all the soil samples, the Er i values for As, Co, Cr, Cu, Ni, Pb and Zn are less than 40 ( Figure 5), except in S6 (road zone), where 7 samples (26%) exceed the value of 40 whose 2 exceed the Er i value of 80. The Er i value of 40 for As is also exceeded in four samples (2 samples in S2 with Er i value of 95, 1 sample in S4 and 1 sample in S5 with an Er i value of 42).
Globally, a low potential ecological risk is thus observed in the studied area but the impregnation with Pb in the road zone is significant, and a large zone is contaminated in such levels of Pb that this zone may be considered as presenting a moderate to considerable ecological risk. A special attention must also be devoted to the As monitoring in S5, where market gardening activities are carried out (artichokes, potatoes, spinach, peas).

Potential health risks
The non-carcinogenic health risks to children living in the vicinity of the study area have been calculated by use of the HQ values for the different exposure pathways of MTE found in topsoils. The HQ values obtained are presented in Table 5 and show a risk in the following order: ingestion > dermal > inhalation. Individually, none of the metals present a HQ superior to 1.0, except Co in the coastal (S1) but above all, Cr in all functional areas (except in S1).
Speciation of chromium has not been conducted during the present study, and it is not possible to know what form of chromium is present (Cr +3 , non-toxic, or Cr +6 , highly toxic).
However, the results found are a very concerning issue for children living in Arzew as Cr +6 is known to induce cancer following oral exposure and chronic low-level skin exposure to Cr +3 or Cr +6 can cause permanent sensitisation that leads to allergic contact dermatitis and both Cr +3 or Cr +6 are known to be respiratory and mucous irritants (Broadway et al. 2010).
Moreover, when considering the combined HI values from the seven metals, high potential non-carcinogenic risk to the local children is observed, with all values exceeding 2.0 and reaching up to 3.7 in the downtown and along the road zone.
Even these results remain uncertain because of the nature of the risk assessment (exposure parameters, duration of exposure, non-inclusion of speciation and bioavailability of metals, but also the absence of determination of MTE in fruits and vegetables grown and consumed in Arzew), and analytical results issued from only one sampling campaign, the health risks posed by the multi-contamination of Arzew soils, appear to seek to force local authorities, at a minimum, to conduct a more extensive study and prevent people of potential risks at playing, working and, growing vegetables in these polluted areas, and at best, to immediate remediation of the contaminated topsoils.

Conclusion
This study investigated for the first time the level of multi-metallic contamination in Arzew, one of the biggest petroleum and industrial platform of Algeria, where more than 100,000 inhabitants coexist with these activities. The distribution patterns of MTE revealed the impact of industrial and human activities on MTE in the topsoils of the study area. Several hotspots contaminated by As, Cr, Pb and Zn have been detected in the road traffic zone but also close to the beach and the harbour. A particular attention should be paid also on the presence of high level of As in the agriculture area. Whereas ecological risk seem to be moderate in almost all the areas, the health risk is high in the whole areas, especially when considering Cr but also the sum of MTE concentrations.
This study reveals thus a multi-metallic contamination over a large surface in Arzew and a hazard to human and ecosystem health, resulting from chaotic expansion of industries and absence of planning. It highlights the urgent need to strengthen petrochemical regulations in order to protect residents from MTE discharges into Arzew' environment.