Evaluation of heavy metal pollution risk in surface sediment of the South Lagoon of Tunis by a sequential extraction procedure

Study area

The South Lagoon of Tunis is the southern basin of the lagoon of Tunis. It is located in the northeast of Tunisia (10°12’ to 10°16’ E and 36°46’ to 36°48’ N). After the restoration programme (1998-2001), the lagoon, covering a surface of about 720 ha, has the shape of as an eclipse stretching in a SW-NE direction. It has a regular depth almost equal to 2.1 m, except in some restricted areas on the east side, where it reaches a maximum of 5 m (Abidi et al. 2019Abidi M., Ben Amor R., Gueddari M. 2019. Sedimentary dynamics of the South Lagoon of Tunis (Tunisia, Mediterranean Sea). Estud. Geol. 75: e086. https://doi.org/10.3989/egeol.43194.487 ).

The lagoon communicates with the Gulf of Tunis through two channels: Channel of Tunis and the Channel of Rades (Fig. 1). The hydrodynamics of the lagoon is mainly controlled by the combination of tide and wind effects (Jouini et al. 2005Jouini Z., Ben Charrada R., Moussa M. 2005. Characteristics of the South Lake of Tunis after restoration. Mar. Life 15: 3-11.). Through the variation of the level of water, the tide controls the opening and the closing of the locks. At rising tide, the seawater enters the lagoon via the Channel of Rades, and it comes out through the Channel of Tunis when the tide is down, imposing a mainly east-west water flow circulation.

Owing to its position, the lagoon is considered an outlet for wastewaters from the neighbouring cities. The South Lagoon of Tunis has a drainage watershed of 4000 ha, more than 1500 ha of which consists of industrial areas (food industries, wholesale, etc.) (Jouini et al. 2005Jouini Z., Ben Charrada R., Moussa M. 2005. Characteristics of the South Lake of Tunis after restoration. Mar. Life 15: 3-11.). The industrial and domestic waters are discharged to the catchment area (Oued Essalaas and the channel of Ben Arous), from where they are transported to the lagoon. In addition, the lagoon receives water from draining of the urban area of Megrine and the runoff of newly-developed areas after the restoration works (Abidi et al. 2019Abidi M., Ben Amor R., Gueddari M. 2019. Sedimentary dynamics of the South Lagoon of Tunis (Tunisia, Mediterranean Sea). Estud. Geol. 75: e086. https://doi.org/10.3989/egeol.43194.487 ). The lagoon receives about 5500 m³/day of untreated industrial wastewater enriched with nutrients and heavy metals (Jouini et al. 2005Jouini Z., Ben Charrada R., Moussa M. 2005. Characteristics of the South Lake of Tunis after restoration. Mar. Life 15: 3-11.).

Sampling and sample handling

Five surface sediments (-3 cm), S-1, S-6, S-8, S-10 and S-12, were extracted from the lagoon in February 2014. The sample locations are shown in Figure 1. The Eh and pH were immediately measured using portable probes (WTW 82862 Weilheim). Each sample was placed in a plastic bag and stored at 4°C for physical and chemical analysis.

In order to separate the fine fraction, the sediment was wet-sieved with bi-distilled water using a 63 µm nylon mesh. Subsequently, it was, oven-dried at 50°C. The fine fraction (<63 µm) was ground in an agate mortar for further analysis. In order to compensate for the grain size and mineralogical effects on the metal variability in different samples, the method of geochemical normalization was applied (Sakan et al. 2009Sakan S.M., Dordevic D.S., Manojlovic D.D. 2009. Trace elements as tracers of environmental pollution in the canal sediments (alluvial formation of the Danube River, Serbia). Environ. Monit. Assess.167: 219-233. https://doi.org/10.1007/s10661-009-1044-0 ). All chemical analyses were carried out on the fine fraction.

Total organic carbon (TOC) in the sediment was measured by means of a Perkin-Elmer PE 2400 CHN analyser. The sub-samples for TOC were decarbonized employing 1 M HCl, rinsed with ultra-pure water and dried at 60°C to remove the carbonate fraction (Froelich 1980Froelich P.N. 1980. Analysis of organic carbon in marine sediments. Limnol. Oceanogr. 25: 242-248. https://doi.org/10.4319/lo.1980.25.3.0564 , Hedges and Stern 1984Hedges J. I., Stern J. K. 1984. Carbon and nitrogen determinations of carbonate-containing solids. Limnol. Oceanogr. 29: 657-663. https://doi.org/10.4319/lo.1984.29.3.0657 , Ennouri et al. 2010Ennouri R., Chouba L., Magni P., Kraiem M.M. 2010. Spatial distribution of trace metals (Cd, Pb, Hg, Cu, Zn, Fe and Mn) and oligo-elements (Mg, Ca, Na and K) in surface sediments of the Gulf of Tunis (Northern Tunisia). Environ. Monit. Assess. 163: 229-239. https://doi.org/10.1007/s10661-009-0829-5 ). A sediment standard supplied by Perkin Elmer (acetanilide, Table 1) was used to calibrate the CHN/S analyser. Calcium carbonate content was assessed by a Bernard calcimeter.

For metallic trace element analysis, 0.5 g of the dried and sieved sediments was placed in a Teflon bomb and digested by adding perchloric (20 mL) and hydrofluoric (10 mL) acid.

The sequential extraction procedure, developed by Tessier et al. (1979)Tessier A., Campbell P.G.C., Bisson M. 1979. Sequential extraction procedure for the speciation of particulate trace metals. Anal. Chem. 51: 844-851. https://doi.org/10.1021/ac50043a017 , was applied in our experiments. It consists of extractions in the following order with associated chemical reagents and conditions from 1 g of sediment:

After each extraction, the separation of the liquid/solid phase was performed by centrifuging the suspension at 3000 rpm for 15 minutes. The supernatant liquid was then separated with a micropipette. The sediment was washed with 10 mL of deionized water and re-centrifuged, while the wash water was discarded. The metal concentrations were determined by flame atomic absorption spectrometry.

Concentrations of metals (Fe, Zn, Pb, Cr, Ni, Cu and Cd) were specified by atomic absorption with flame using the Perkin Elmer Analyst 100 Spectrometer. Calibration standards of AAS were prepared by serial dilution and verified against standard reference materials.

To show the quality, accuracy and repeatability of the extraction, certified-reference sediment samples and blanks were analysed. To validate the analysis results, we mineralized certified sediment (IAEA-405 reference material, trace elements and methyl mercury in estuarine sediment) at predefined concentrations (Table 1).

The correlations of the studied elements were calculated using the XLSTAT (2013) software.

Various indices have been suggested to quantify the degree of metal contamination in sediments and to assess the potential health risk (Müller 1979Müller G. 1979. Schwermetalle in den sedimenten des RheinsVeränderungen seit 1971. Umschau in Wissenschaft und Technik 79: 778-783., Salomons and Förstner 1984Salomons W. and Förstner U. 1984. Metals in the Hydrocycle. Springer-Verlag, Berlin, 349 pp. https://doi.org/10.1007/978-3-642-69325-0 , Ioannides 2015Ioannides K., Stamoulis K., Papachristodoulou C., et al. 2015. Distribution of heavy metals in sediment cores of Lake Pamvotis (Greece): a pollution and potential risk assessment. Environ. Monit. Assess. 187: 4209. https://doi.org/10.1007/s10661-014-4209-4 ). In the present study, the following factors were taken into consideration:

The geoaccumulation index (Igeo) was employed to assess the metal contamination in sediments, compared with a background geochemical reference (Müller 1969Müller G. 1969. Index of geoaccumulation in sediments of the Rhine River. Geol. J. 2: 109-118.). Igeo was calculated applying the following equation:

where C_n is the measured concentration of the metal (n) in the sample and B_n denotes the geochemical background concentration of the metal (n). The factor 1.5 was utilized to minimize the effect of possible variations in the sediments. Müller (1969)Müller G. 1969. Index of geoaccumulation in sediments of the Rhine River. Geol. J. 2: 109-118. distinguished seven classes of the geoaccumulation index:

The enrichment factor (EF) is commonly used as a means of identifying and quantifying the anthropogenic impact on sediments for a metal (m). It was calculated according to Salomons and Förstner (1984)Salomons W. and Förstner U. 1984. Metals in the Hydrocycle. Springer-Verlag, Berlin, 349 pp. https://doi.org/10.1007/978-3-642-69325-0 as follows:

where ${\left[\raisebox{1ex}{$Cm$}\!\left/ \!\raisebox{-1ex}{$Cr$}\right.\right]}_{sediment}$ and ${\left[\raisebox{1ex}{$Cm$}\!\left/ \!\raisebox{-1ex}{$Cr$}\right.\right]}_{Background}$ are the concentration ratios of element m to the reference element R in the sample and background, respectively. EF values may be interpreted as suggested by Birch and Davies (2003)Birch G.F, Davies K.I. 2003. A scheme for assessing human impact and sediment quality in coastal waterways. In: Woodroffe C.D., Furness R.A. (eds), Proceedings of the coastal GIS conference, Wollongong, NSW, 7-8 July, 2003: pp. 371-380. Australia: University of Wollongong.:

The EF, Igeo and C_f for each heavy metal were computed as a function of Fe content, which is a conservative element considered as the detrital fraction of sediment and compared with the amount of Fe in sediments during the preindustrial period. It has been successfully used to normalize metal contaminants in several studies, as revealed by Ioannides (2015)Ioannides K., Stamoulis K., Papachristodoulou C., et al. 2015. Distribution of heavy metals in sediment cores of Lake Pamvotis (Greece): a pollution and potential risk assessment. Environ. Monit. Assess. 187: 4209. https://doi.org/10.1007/s10661-014-4209-4 .

The contamination approximation is defined according to the natural content “background”. To avoid a bad estimation of the pollution indicators, it is necessary to use a local background, i.e. a local geochemical background. In our study, we used a deep sediment (at -8 m), which corresponds to a preindustrial level of the lagoon (Ouertani and Yahyaoui 2019Ouertani N., Yahyaoui S. 2019. Holocene Paleoclimatic Variation Inferred from Study of Sediments in the Gulf of Tunis (North Africa). Book chapter: Patterns and Mechanisms of Climate, Paleoclimate and Paleoenvironmental Changes from Low-Latitude Regions, Adv. Sci. Technol. Innov. pp 37-40. https://doi.org/10.1007/978-3-030-01599-2_9 ).

Sediment pollution index (SPI): This index is generally used for an overall assessment of sediment quality with respect to heavy metal concentrations along with a proper consideration of the relative metal toxicity. It is the linear sum of the metal enrichment factors. It takes into account metal toxicity weights. The metal toxicity weights are based on the relative toxicity of different metals. Weight 1 was assigned to Cr and Zn, the less toxic metals; 2 to Ni and Cu; 5 to Pb; and 300 to Cd (Rubio et al. 2000Rubio B., Nombela M.A., Vilas F. 2000. Geochemistry of major and trace elements in sediments of the Ría de Vigo (NW Spain): An assessment of metal pollution. Mar. Poll. Bull. 40: 968-980. https://doi.org/10.1016/S0025-326X(00)00039-4 , Singh et al. 2002Singh M., Müller G., Singh I.B. 2002. Heavy Metals in freshly deposited stream sediments of rivers associated with urbanisation of the Ganga Plain, India. Water Air Soil Pollut. 141: 35-54. https://doi.org/10.1023/A:1021339917643 ).

The SPI can be expressed as follows (Rubio et al. 2000Rubio B., Nombela M.A., Vilas F. 2000. Geochemistry of major and trace elements in sediments of the Ría de Vigo (NW Spain): An assessment of metal pollution. Mar. Poll. Bull. 40: 968-980. https://doi.org/10.1016/S0025-326X(00)00039-4 , Singh et al. 2002Singh M., Müller G., Singh I.B. 2002. Heavy Metals in freshly deposited stream sediments of rivers associated with urbanisation of the Ganga Plain, India. Water Air Soil Pollut. 141: 35-54. https://doi.org/10.1023/A:1021339917643 ):

where EF is the enrichment factor of each metal and W designates the toxicity weight of each metal.

Based on the above calculation, these classes are obtained:

General characteristics of the sediment

Data related to grain size, TOC, CaCO₃, pH and Eh are presented in Table 2. Surface sediments of the South Lagoon varied slightly in texture. The lagoon is mantled in a silty fraction (95%) and a sandy fraction (5%). The restoration project, especially the establishment of two groups of inlets gates of the channel of Rades and Tunis (recalibrated during the restoration management), imposed an east-west water flow direction. This current controls local sediment sorting (Abidi et al. 2019Abidi M., Ben Amor R., Gueddari M. 2019. Sedimentary dynamics of the South Lagoon of Tunis (Tunisia, Mediterranean Sea). Estud. Geol. 75: e086. https://doi.org/10.3989/egeol.43194.487 ).

The eastern side of the lagoon, near the channel of Rades, has the highest percentage of sand fraction. The central and western sides of the lagoon are covered by fine sand and deposited in relatively calm hydrodynamic conditions. In fact, the sediment dynamics is controlled by the lagoon water currents inducing east-west grain-size sorting.

Eh values varied between -27 and -134mV, reflecting anoxic conditions in the lagoon sediments. Overall, Eh depended on sediment characteristics such as the percentage of coarse fraction, TOC content and bioturbation (Oueslati et al. 2010Oueslati W., Added A., Abdeljaoued S. 2010. Geochemical and statistical approaches to evaluation of metal contamination in a changed sedimentary environment: Ghar El Melh Lagoon, Tunisia. Chem. Speciat. Bioavailab. 22: 227-240. https://doi.org/10.3184/095422910X12893267432461 ). Obviously, the Eh was higher on the eastern side of the lagoon, which can be explained by the diffusion of more oxygenated waters coming from the Gulf of Tunis and by the porosity of the sandy sediments near the channel of Rades. pH values did not change much in the lagoon sediments.

The highest Eh values recorded near the channel of Rades resulted from bioirrigation, waves and currents (Oueslati et al. 2018Oueslati W., Helali M.A., Mensi I., Bayaoui M., Touati H., Khadraoui A., Zaabooub N., Added A., Aleya L. 2018. How useful are geochemical and mineralogical indicators in assessing trace metal contamination and bioavailability in a post-restoration Mediterranean lagoon? Environ. Sci. Pollut. Res. 25: 25045-25059. https://doi.org/10.1007/s11356-018-2575-0 ). Therefore, the east side of the lagoon seems to be more affected by the incoming marine flux from the Gulf.

As shown in Table 2, the calcium carbonate content ranged from 26.91% to 32.55%, showing a decreasing trend from east to west. The highest CaCO₃ content was recorded in the samples with a high coarse fraction. Obviously, the latter was dominated by complete and fragments of shells, suggesting that carbonates in the lagoon have a primarily biogenic origin (Abidi et al. 2019Abidi M., Ben Amor R., Gueddari M. 2019. Sedimentary dynamics of the South Lagoon of Tunis (Tunisia, Mediterranean Sea). Estud. Geol. 75: e086. https://doi.org/10.3989/egeol.43194.487 ). The concentration of the TOC varied between 1% (0.48 µg/g) and 1.69% (1.84 µg/g), with an increasing trend in the west.

Metal concentrations

The metal concentrations in the lagoon surface sediment and the regional background values are listed in Table 3.

Ranges of metal concentrations in sediments were 38589.3 to 43705.74 µg/g for Fe with an average of 41727.19 µg/g, 125.8 to 188.29 µg/g for Zn with an average of 148.52 µg/g, 69.33 to 147.61 µg/g for Pb with an average of 102.22 µg/g, 22.68 to 28.91 µg/g for Cu with an average of 27.28 µg/g, 89.83 to 118.29 µg/g for Cr with an average of 99.84 µg/g, 1.18 to 1.9 µg/g for Cd with an average of 1.51 µg/g and 53.49 to 86.59 µg/g for Ni with an average of 71.81 µg/g.

Fe concentrations showed no significant variations among sites. The amount of Zn, Cu, Cr and Cd was higher on the eastern side. This increase was more pronounced for Pb and Ni.

The highest metal concentrations were found on the east side of the lagoon as a result ofo metal-containing wastewater discharges from the urban complex in the watershed. The highest mean values of Zn, Pb, Cr and Ni were found at site S-12, and the highest mean values of Cu were found at site S-10. The sediments from the inner channel of Rades (S-1) had the lowest mean metal concentrations.

To assess the state of contamination of metals in surface sediments in the South Lagoon of Tunis, total mean metal concentrations were compared with those in the North Lagoon of Tunis and those before the restoration project (Table 3). The mean concentrations of total trace metals of our sediments were compared with those of other areas. Unlike the mean concentrations of Ni, which remained high, those of Zn (148.5 µg/g), Pb (102.2 µg/g), Cu (27.3 µg/g), Cr (99.8 µg/g) and Cd (1.5 µg/g) are lower than those measured before the restoration project, as reported by Ben Souissi et al. (1999)Ben Souissi J., Zaouali J., Aouij S., et al. 1999. Teneurs en metaux traces des sediments de surface du Lac Sud de Tunis avant restoration. IAEA-SM-354/14:13-18.. The overall average concentrations of metals in surface sediments of the lagoon were below or close to the average contents recorded before the restoration project.

The comparison of metal concentrations in the South Lagoon of Tunis with those in other Mediterranean lagoons reveals that the Tunisian lagoons (South Lagoon of Tunis, Ghar el Melh and Korba) (Table 4) have a generally high level of metals because of the rudimentary sanitary network and the overlap of industrial and domestic wastewater discharged into the lagoons. However, the Ni content in the South Lagoon of Tunis is lower than that in the Thau lagoon and the Homa Lagoon. Otherwise, concentrations of TOC are quite similar.

Granulometry, physico-chemical parameters and organic matter content are important factors that influence the geochemical behaviours of heavy metals in sediments. In order to reveal possible associations between these variables, Pearson’s correlation was performed. The correlation coefficient matrix is presented in Table 5.

A significant positive correlation was observed for Zn, Pb, Cr, Cu, and Ni (taken in pairs), indicating that these metals are associated with each other and may have similar natural or anthropogenic sources and are governed by the same process. This strong correlation between Cu, Pb, Zn, Ni and Cr was also reported in the literature for other urbanized and polluted areas (Ruiz 2001Ruiz F. 2001. Trace metals in estuarine sediments from the southwestern Spanish coast. Mar. Poll. Bull. 42: 482-490. https://doi.org/10.1016/S0025-326X(00)00192-2 , Spencer 2002Spencer K.L. 2002. Spatial variability of metals in the inter-tidal sediments of the Medway estuary, Kent, UK. Mar. Poll. Bull. 44: 933-944. https://doi.org/10.1016/S0025-326X(02)00129-7 , Muniz et al. 2003Muniz P., Danulat E., Yannicelli B., et al. 2003. Assessment of contamination by heavy metals and petroleum hydrocarbons in sediments of Montevideo harbour (Uruguay). Environ. Int. 1096: 1-10., Baptista Neto et al. 2006Baptista Neto J.A., Gingele F.X., Leipe T., Brehme I. 2006. Spatial distribution of heavy metals in surficial sediments from Guanabara Bay: Rio de Janeiro, Brazil. Environ. Geol. 49: 1051-1063. https://doi.org/10.1007/s00254-005-0149-1 ). However, Cd showed a low correlation, proving that it has different metal sources. The strong positive correlations between metals and TOC suggest that Pb, Cu, Cr and Ni contents increase with the rise of the amount of organic matter in the sediments. As demonstrated by Horowitz and Elric 1987Horowitz A.J., Elric K.A. 1987. The relation of stream sediment surface area, grain size and composition to trace element chemistry. Appl. Geochem. 2: 437-451. https://doi.org/10.1016/0883-2927(87)90027-8 , Book and Moore 1988Book E., Moore J.N. 1988. Particle-size and chemical control of As, Cd, Cu, Fe, Mn, Ni, Pb and Zn in bed sediment from the Clark Fork river, Montana (U.S.A.). Sci. Total Environ. 76: 247-266.https://doi.org/10.1016/0048-9697(88)90111-8 , Stone and Droppo 1996Stone M., Droppo I.G. 1996. Distribution of lead, copper and zinc in size-fractionated river bed sediment in two agricultural catchments of southern Ontario, Canada. Environ. Pollut. 93: 353-362. https://doi.org/10.1016/S0269-7491(96)00038-3 , Murray et al. 1999Murray K.S., Cauvet D., Lybeer M., Thomas J.C. 1999. Particle size and chemical control of heavy metals in bed sediment from the Rouge river, Southeast Michigan. Environ. Sci. Technol. 15: 474-480 https://doi.org/10.1021/es9807946 and Li et al. 2001Li X., Shen Z., Wai O.W., Li Y.S. 2001. Chemical forms of Pb, Zn and Cu in the sediment profiles of the Pearl River Estuary. Mar. Poll. Bull. 42: 215-223. https://doi.org/10.1016/S0025-326X(00)00145-4 , organic matter is an important carrier of metallic trace elements in sediments.

Chemical speciation

In sediments, metallic trace elements exist in several different forms and are associated with a range of components (Cottenie et al. 1979Cottenie A., Camerlynck R., Verloo M., Dhaese A. 1979. Fractionation and determination of trace elements in plants, soils and sediments. Pure Appl. Chem. 52: 45-53. https://doi.org/10.1351/pac198052010045 ). It is recognized that information about the physico-chemical forms of each element is necessary to understand their environmental behaviour (mobility, pathways and bioavailability) (Bernhard et al. 1986Bernhard M., Brinckman F.E., Sadler P.J. 1986. The importance of chemical “speciation” in environmental processes. In: Bernhard M., Brinckman F.E. and Sadler P.J. (eds), Report of the Dahlem Workshop, Life Sciences Research Report 33, Berlin, September 2-7, 1984, Springer-Verlag, Berlin, Heidelberg, 1986, 762 pp. https://doi.org/10.1016/0048-9697(87)90182-3 ).

In this study, the distributions of different fractions of heavy metals in surface sediments of the South Lagoon of Tunis are presented in Figure 2.

In the sediment of the South Lagoon of Tunis, Fe was almost entirely extracted in the residual phase (more than 90%) followed by the organic matter-sulphide phase. The other phases were negligible or even absent. Chemical speciation of lead revealed its predominant association with the Fe and Mn oxide fractions and to a lesser extent with organic matter and/or sulphides. The residual fraction also played a less important role in lead sorption. More than 40% of Zn was concentrated in the Fe and Mn oxide fraction. From S-1 to S-12, the Zn percentage associated with a residual fraction (F5) decreased slightly at the expense of the organic matter fraction.

The Cu distribution showed a marked association with the organic phase, with contents as high as 65% at S8, at the expense of the carbonate phase and the phase linked to the oxyhydroxide of Fe and Mn. The greatest proportion of Cu was mainly related to the organic matter-sulphide fraction, particularly at S-8, followed by the residual fraction. The exchangeable fraction was relatively negligible. Cr and Ni were combined preferentially with the residual fraction and, to a lesser extent, with organic matter and/or sulphides and Fe/Mn oxides. The other phases (the exchangeable fraction and the fraction linked to carbonates) were negligible. Cd was essentially associated with the Fe/Mn oxy-hydroxides fraction. Cadmium showed a high distribution linked to carbonates in comparison with other metals. Furthermore, the organic matter-sulphide fraction was negligible in comparison with the other fractions (exchangeable fraction, Fe/Mn oxy-hydroxides fraction and residual fraction), except at S-8.

Generally, the highest proportion of extractable metals is found in sediments with a high clay and silt content. Measuring the total content of metals is a fundamental method for sediment quality assessment. However, to further understand the potential mobility, bioavailability and toxicity of heavy metals in sediments, metal fractionation occurring in different geochemical forms is of crucial importance. Trace elements in the different geochemical phases are characterized by different mobility, migration ability and chemical behaviour. Therefore, the sequential extraction procedure was proposed to obtain information about the ways associating metal with sediments (Wang et al. 2015Wang Z., Wang Y., Chen L., et al. 2015. Assessment of metal contamination in coastal sediments of the Maluan Bay (China) using geochemical indices and multivariate statistical. Mar. Poll. Bull. 99: 43-53. https://doi.org/10.1016/j.marpolbul.2015.07.064 ).

Through chemical fractionation, it was possible to determine that metals bound mainly to the residual fraction indicate no environmental risk, and that those predominantly associated with the mobile fractions (exchangeable fraction + carbonates + iron-manganese oxyhydroxides + organic matter-sulphides) show anthropogenic influences.

The “exchangeable fraction + carbonates fraction” forms the acid-soluble fraction. The first two fractions have the weakest bond to sediments and become, which makes them the most mobile. The higher the metal content is in this fraction, the easier their extraction will be and the more bioavailable they become; hence they involve a high risk to the environment.

Among the studied metals, the relative proportions of metals in the acid-soluble fraction were generally low (Fe 29.7%, Pb 20.6%, Zn 6.7%, Cr 41.8% and Ni 13.8%), except for Cd. The contents of acid-soluble Cd on average (41.8%) were higher than those of other Cd fractions. It is distributed more as iron and manganese oxide forms (reducible forms) and as carbonates. Obviously, metal removal required strong acid conditions; hence their poor association with the carbonate and exchangeable phases. Exchangeable species play a minor role in metal fixation in the lagoon sediments. They are usually related to the adsorbed metals on the sediment surface that can be easily remobilized into the lagoon water.

Fe/Mn oxyhydroxides, or the reducible fraction, was the most abundant fraction. Zn (49.6%), Pb (54%), Cr (31%), Cd (39.3%) and Ni (25%) were largely associated with the Fe/Mn oxyhydroxides (reducible forms). They are excellent trace element scavengers (Jenne 1968Jenne E.A. 1968. Trace inorganics in water. Adv. Chem. Ser. 73: 337-387. https://doi.org/10.1021/ba-1968-0073.ch021 , Oakley et al. 1981Oakley S.M., Nelson P.O., Williamson K.J. 1981. Model of trace-metal partitioning in marine sediments. Environ. Sci. Technol. 15: 474-480. https://doi.org/10.1021/es00086a015 ). However, metal-Fe/Mn oxyhydroxides can be mobilized under anoxic conditions (Tessier and Campbell 1987Tessier A., Campbell P.G.C. 1987. Partitioning of trace metals in sediments: relationship with bioavailability. Hydrobiologia 149: 43-52. https://doi.org/10.1007/BF00048645 ).

The percentage of Cu bound to the organic matter fraction (the oxidizable fraction) was more higher than that of other metals bound to this fraction, with a range of 31%. According to Li et al. (2001)Li X., Shen Z., Wai O.W., Li Y.S. 2001. Chemical forms of Pb, Zn and Cu in the sediment profiles of the Pearl River Estuary. Mar. Poll. Bull. 42: 215-223. https://doi.org/10.1016/S0025-326X(00)00145-4 , copper tends to form stable complexes with organic matter.

Under oxidizing conditions, metals present in both natural organic matter (due to complexation) and living organisms (as a result of bioaccumulation of metals) may be remobilized into the aquatic environment. The results obtained show that most of the Ni and Fe in all the sediments was strongly retained in the residual fraction. Their average percentage in this fraction was 37.9% and 34.8%, respectively. These facts indicate that these metals are strongly bound in the crystal lattices of minerals and, consequently, they have relatively low mobility, bioavailability and toxicity (Wang et al. 2015Wang Z., Wang Y., Chen L., et al. 2015. Assessment of metal contamination in coastal sediments of the Maluan Bay (China) using geochemical indices and multivariate statistical. Mar. Poll. Bull. 99: 43-53. https://doi.org/10.1016/j.marpolbul.2015.07.064 ). Copper exists mostly as oxidizable species and also as iron and manganese oxide forms (reducible forms).

Contamination assessment

In order to evaluate sediment pollution in the lagoon, the geo-accumulation index (Igeo) and the EF were calculated using total concentrations of metals (Table 6).

The results of calculating the geo-accumulation index in the sediments of the lagoon show that Zn and Cr did not reflect any pollution, with an Igeo of between -0.69 and -0.11. For Fe, Pb, Cu and Cd, the Igeo indicated slight pollution. The lagoon sediments showed strong Ni pollution. The highest values were recorded west of S-12. The Igeo results show that lagoon sediment pollution by Ni was the most acute, while it was moderate for Cu. The Igeo calculated for Fe, Pb and Cd showed uncontaminated to moderately contaminated sediments, while those calculated for the other metals (Zn and Cr) were uncontaminated. EF values for Zn, Cr, Fe, Cd and Pb ranged from 0 to 2.5, indicating moderate contamination. However, for Ni and Cu, the sediments of the lagoon were considerably contaminated. The highest EF values characterized C12 sediments. The EF values showed no enrichment of Zn and Cr to minor enrichment of Pb, Cu, Cd and Ni. The highest values were recorded at S12.

The sediment pollution index values demonstrate that the sediments of the lagoon are slightly polluted (SPI <5), except for the S-12 sediments, which have strong indices of sediment pollution.

However, only the mobilizable fractions (i.e. exchangeable species, carbonate bound metals, Fe/Mn oxides species and organic matter-bound metals) in the sediments showed a real potential risk of water contamination caused by lagoon sediment. The remobilization of metals from sediment into the water column is influenced by pH, chemical forms of the trace elements and the physico-chemical characteristics of the water column (Ikem et al. 2003Ikem A., Egiebor O.N., Nyavor K. 2003. Heavy elements in water, fish and sediment from Tuskegee Lake, southern USA. Water Air Soil Pollut. 149: 51-75. https://doi.org/10.1023/A:1025694315763 ). To estimate the risk of remobilization of metal from the sediment to the water column and to the lagoon biota, the individual contamination factor (ICF) and the global contamination factor (GCF) were calculated as shown by Ikem et al. (2003)Ikem A., Egiebor O.N., Nyavor K. 2003. Heavy elements in water, fish and sediment from Tuskegee Lake, southern USA. Water Air Soil Pollut. 149: 51-75. https://doi.org/10.1023/A:1025694315763 and Zhao et al. (2012Zhao S., Feng C., Yang Y., Niu J., Shen Z. 2012. Risk assessment of sedimentary metals in the Yangtze Estuary: New evidence of the relationships between two typical index methods. J. Hazard. Mater. 241. https://doi.org/10.1016/j.jhazmat.2012.09.023 ) (Tables 7 and 8). The factor calculation shows that the lagoon sediments had a high ICF for Pb (ICF: 6-14) and Cd (ICF: 4.1 to 7.9), a moderate ICF for Ni (1.3 to 2), Cu (1.3 to 4.2), Zn (1.8 to 5.9) and Fe (1.5 to 2.3) and a low ICF for Cr (0.7 to 4.3). Surface sediments showed a considerable to high global risk with a GCF varying from 21.3 to 40.