Contamination of Surface Sediments by Heavy Metals in Ebrié Lagoon (Abidjan, Ivory Coast)

INTRODUCTION

Several human activities (agriculture, industry, transport, fishing, port, etc.) are developed around water points because water still the most important natural resource (Tuo et al., 2012). Thus, estuarine and coastal areas exhibit a wide array of anthropogenic inputs that might compromise their ecological integrity, with their rapid population growth and uncontrolled development in many coastal regions worldwide. Waters of estuarine environment are in continuous interactions with marine waters. These interactions influence the heavy metal speciation into water, sediments, plants and aquatic organism (Ip et al., 2007). Metals are neither created nor destroyed by human, they are just transformed for several needs. Except for contamination, metals also occur in low concentrations in aquatic system (Adefemi et al., 2007; Ndimele et al., 2011). Different pollutants such as heavy metals, from domestic, industrial, port activities can be easily stabilized and mobilized in the sediments (Siddique and Aktar, 2012). These heavy loads of pollutants are responsible for the availability of heavy metals in coastal areas. In Ivory Coast, more than 80% of the industries and factories are located on the bank of the bays or very close to the Ebrié lagoon system and most of them don’t have any waste treatment facilities. Consequently, they discharge the untreated wastes into the nearest water bodies which finally reach into the Ebrié lagoon through different canal system. Thus, the urban area of this important lagoon is increasingly contaminated with toxic substances, threatening all forms of life (Al-Hashem and Brain, 2009; Sarma, 2011; Orlu and Gabriel, 2011; Tuo et al., 2012). Heavy metals are known to be accumulated by marine organisms through a variety of pathways that include respiration, adsorption and ingestion (Zhou et al., 2001). Heavy metals can be accumulated in waters, sediment, plants and aquatic organisms in estuarine ecosystems. Thus, due to their toxicity, contamination of these ecosystems by heavy metals is of major concern for all forms of life. Pollution by heavy metals is a threat to human life and the total environment (Igwe and Abia, 2006). Heavy metals are an important class of pollutants in the aquatic environment. Monitoring the heavy metals concentrations in sediments is important for toxicological studies in aquatic environments. Due to their long residence time, heavy metals concentrations in sediments remain a vital source of information regarding their sources, speciation and degree of contamination or pollution (Adefemi et al., 2007; Madkour et al., 2012). This is for the fact that sedimentation has been regarded as one of the most important fluxes in aquatic systems (Asaolu et al., 1997). In natural waters, metals distribution processes are controlled by a dynamic set of physical-chemical interactions and equilibria and their solubilities are controlled by pH, concentration, type of metal species, organic ligands, the oxidation state of mineral components and the redox environment of the aquatic system (Huang and Lin, 2003; Jeon et al., 2003; Millward and Liu, 2003; Warren et al., 2005; Grosbois et al., 2006).

Some heavy metals such as Pb, Cu, Zn, etc., have been shown in some previous investigations to occur at a significant level in the waters, sediments and aquatic organisms (Metongo and Gbocho, 2007; Kone et al., 2008; Marcellin et al., 2009; Yao et al., 2009; Tuo et al., 2012). However, there are no reports available regarding large scale assessment of organic matter and heavy metals contents in sediments of the estuarine area of the Ebrié lagoon.

This study represents baseline information regarding anthropogenic impacts on the Ebrié lagoon by the assessment of organic matter levels and concentrations of some heavy metals (Pb, Cd, Ni, Mn and Zn) in the estuarine part of the Ebrié lagoon. The degree of heavy metals contamination of the area was evaluated by calculating the Contamination Index (CI) and the Mean Contamination Index (MCI) for the different locations. Hence, it will be useful for the management and sustainable development of this socio-economical area.

MATERIALS AND METHODS

Study area: The Ebrié lagoon has an area of 566 km² and stretches on 125 km along the coast of Ivory Coast, between 3°40' and 4°50' West, at latitude 5°50' North (Table 1). Its volume is estimated at 2.5x0⁹ m³, with an average depth of 4.8 m and a few pits near Abidjan (Bay Abou-Abou) that exceed 20 m (Dufour, 1982).

Table 1:	Names and No. and GPS coordinate of the sampling stations in the Ebrié lagoon

This lagoon communicates with the Atlantic Ocean by the Vridi channel, drilled in 1951, for the building of Abidjan port, the most important in West Africa.

Collection of sediment samples: For this study, 144 surface sediment (approximately 0-2 cm layer) samples have been collected in Ebrié Lagoon from February 2008 to December 2009, using a stainless steel Van Been grab, at a mean distance of 25-50 m from domestic and/or industrial inputs. The exact position of each sampling site was recorded using Global Positioning System (GPS). The sampling sites were selected on the accessibility and the inputs nature. Samples were collected from eighteen stations located in five urban bays (Banco, Cocody, Biétri, Marcory and Milliardaires) and in a rural one (Bingerville chosen as the reference), close to Bingerville city and far from urban inputs (Table 1). Lagoon surface sediments were chosen for this study as this layer controls the exchange of metals between sediments and water (El Nemr et al., 2006; Praveena et al., 2008). All sampling equipment and laboratory apparatus were acid washed and rinsed thoroughly first with tapwater and then with distilled water to ensure any traces of cleaning reagent were removed before the sampling and/or analysis. Finally, they were dried at room temperature (25°C) and store in a clean place. The sediments were kept cool in an icebox during transportation to the laboratory.

Physicochemical characterization: The physicochemical parameters such as pH, salinity, conductivity and temperature were measured in situ with a multi-parameter (type WTW pH/340i).

Analysis of sediment samples: Organic matter and heavy metals were carried out in duplicates and the average data were considered. Each sediment sample was mixed well and then 30 g was taken into a glass dish and dried at 105°C in a Gallen Kamp oven for 24 h. Samples were then homogenize using pestle and Mortar, passed through a 2 mm and <63 μm mesh screens, respectively and stored in polyethylene bags. The Walkley and Black (1934) method was used to determine the organic carbon content. In this procedure, organic carbon was determined using sulphuric acid and aqueous potassium dichromate (K₂Cr₂O₇) mixture. After complete oxidation from the heat of solution and external heating, the unused or residual K₂Cr₂O₇(in oxidation) was titrated against ferrous ammonium sulphate. The used K₂Cr₂O₇ (the difference between added and residual K₂Cr₂O₇), gave a measure of organic carbon content of the sediment. For heavy metals, about 0.3 g of the prepared sediment sample was weighed accurately to 0.01 g accuracy and completely digested in a Teflon bomb by using 1 mL of acid mixture (HNO₃: HCl; 1/3, v/v ) and 4 mL of fluorhydric acid, respectively. Acid were slowly added to the dried sample and left overnight before heating. Samples were heated for 2 h 30 min on hot plate at temperature of approximately 120°C. After cooling, the digested samples were filtered and kept in plastic bottles and the solution was justified to volume 25 mL with distilled water. The concentrations of the elements were determined by Atomic Absorption Spectrophotometer (AAS) (Perkin-Elmer, 3030 Model) with air/acetylene flame, at specific wavelengths and after preparation of calibration standards.

The degree of metal pollution was assessed using the Contamination Index (CI) and the Mean Contamination Index (MCI). In the absence of "background levels" (natural content) without anthropogenic inputs, several authors have proposed the calculation of the index of contamination to establish the degree of contamination of a site compared to overall pollution in the whole region (Chafika et al., 2001; Kaimoussi et al., 2002), according to the equation:

where, Mx = metal mean value at station x (in mg kg^-1); M = y mean level for all stations (in mg kg^-1). Mean Contamination Index (MCI) was determined using the following equation in order to classify the stations:

where, n is the total number of metals analyzed (for the present study, n= 5).

Statistical analysis: All statistical analyzes (Mean, standard deviation, Two-way ANOVA, etc.) were performed with STATISTICA software (2005, 7.1 Version), a complete integrated statistical data analysis. Factorial analysis was necessary to capture the main effects and possible interactions between locations and seasons. Correlation circle and matrix were performed with ADE-4 to capture the main differences between locations. Significant treatment means were compared using least significant difference at 5% probability level.

RESULTS AND DISCUSSION

Physicochemical parameters: The estuarine area of the Ebrié lagoon is the main receptacle of domestic, industrial and port effluents, torrents of tropical rainfall from Abidjan city. Almost all of these effluents are untreated, leading to the pollution of this ecosystem by several hazardous pollutants such as heavy metals. Some immediately measurable physicochemical parameters were assessed and their mean values and standards deviations were presented in Table 2.

The highest mean value for temperature was recorded at S1 (26.25±5.44°C) while the least (21.70±5.72°C) was recorded in sediments collected at BNETD. Significant differences (p<0.05) were observed between sites. For temperature, significant differences (p<0.05) were observed between seasons.

The highest mean value for pH was recorded at Sebroko (7.90±0.23) while the least (7.04±0.04) was recorded at S3. Sites were significantly different (p<0.01) for pH due to their individual characteristics (depth, renewal rate, hydrodynamism, chemical characteristics of domestic and/or industrial inputs, etc.) (Arfi and Guiral, 1994). For the present study, pH levels were appropriate for the support of aquatic life (4.50-10.00) (Yao et al., 2009).


Fig. 1:	Variations of heavy metal concentrations in Ebrié Lagoon sediments

Table 2:	Analysis of some limnological parameters in the Ebrié Lagoon sediments

Values are Mean±SD

The highest mean values for salinity and conductivity were recorded at Sebroko (2.11±0.43%, 30.90±11.59 mS cm^-1) while the least (0.35±0.48%, 8.62±5.21 mS cm^-1) were recorded at R, farther from Vridi channel (Fig. 1). For salinity and conductivity, sites were not significantly different (p>0.05). At the opposite, values of salinity and conductivity have significantly varied with seasons (p<0.01). Indeed, the highest levels of salinity were observed on dry seasons, when marine waters dilute the lagoon waters. In contrast, on rainy seasons, fresh waters predominate which leads to the decrease of salinity and conductivity.

The highest mean value of organic matter was observed at SIVOA (67.37±14.07%) while the least was observed at R (8.05±7.32%). During the whole period of the study, highest levels of organic matter have been observed at Biétri bay (Fig. 1). Organic matter accumulation observed in sediments collected at urban bays were the result of organic inputs through untreated domestic, industrial, etc., effluents that often contain highest loads of organic components. No significant difference was observed between locations and seasons for organic matter during the present study, due to continuous inputs of organic loads in each site whatever their location and the seasons.

Heavy metals
Spatial variations: Sediments represent a potential source of contaminants to the overlying water and hence can influence water quality. The natural release of sediment contaminants is controlled by their dissolution into the sediment pore waters. Diffusion of these contaminants to the water column will occur if the pore water concentration exceeds that of the overlying water. Indeed, due to interactions between surface sediments and water at water-sediment interface, knowledge about heavy metals in sediments is a necessity in toxicological studies. Spatial variations of heavy metals are presented in Fig. 1.

Highest value of Pb (536.09±217.70 mg kg^-1) was recorded at CARENA and the least (1.45±5.45 mg kg^-1) at R. For Cd, highest value (8.36±12.92 mg kg^-1) was recorded at Stade FHB while the least (0.05 mg kg^-1) was observed at S3 and R. Highest value of Ni (103.25±164.68 mg kg^-1) was observed at Bidet and the least (0.54±2.36 mg kg^-1) at R. The maximum value for Mn (281.0±405.0 mg kg^-1) was observed at S2 and the minimum value (18.28±5.23 mg kg^-1) at R while the highest value for Zn (710.12±36.30 mg kg^-1) was recorded at Unilever and the lowest value (4.68±6.13 mg kg^-1) at R.

In general, levels of all heavy metals observed were lower and found save for aquatic life in sediments collected at R, close to Bingerville city. The lowest levels of heavy metals observed at R were due to its location, farther from Abidjan. Also, highest values of heavy metals often observed in sediments at Bay Milliardaires may be due to several construction jobs made in recent years, without any sanitation system. At Bay Marcory, highest levels for Pb, Zn, Ni and Cd were recorded at Biafra. Highest levels of Pb and Zn were observed at S2 and at Biafra, respectively. These highest values observed at Biafra for these metals might be the result of several activities (car repair shops, car wash, domestic and industrial activities, etc.) conducted without control over the station’s watershed. Highest levels of zinc observed could affect reproductive capacity of some organisms such as fishes, oysters and larval growth (Osman et al., 2009). Except for manganese levels observed by Ochieng et al. (2007) in Lakes of Kenya, levels observed for the present study were higher than those observed elsewhere (Table 5) (Ochieng et al., 2007; Issola et al., 2009; Marcellin et al., 2009; Ndimele et al., 2011; Madkour et al., 2012). Pb and Zn concentrations observed by Marcellin et al. (2009) were lower than those observed during the present study, leading to the fact that these elements were still introduced in the lagoon through anthropogenic inputs (Marcellin et al., 2009). Except for Mn, significant differences (p<0.0001) were observed between sites. Spatial variations have revealed that hazardous pollutants sources in Ebrié lagoon waters depend on the location of each site (Fig. 1). Except for Mn, for which higher concentrations (90-1471 mg kg^-1) were observed in El-Hamrawein Harbour (Red Sea, Egypt) sediments, the concentrations of the other metals (Pb, Cd, Ni and Zn) observed in the Ebrié lagoon were higher than those observed elsewhere (Table 5). To appreciate the degree of pollution of the studied sediments, some guideline values established according to the different effects observed on living organisms following toxicology tests were considered (Table 3 and 4). Table 3 and 4 represent five increasing categories of observable effects on aquatic organisms. Levels of heavy metals observed in the Ebrié lagoon where then compared with the sediments guidelines and those observed by other authors (Table 5). The comparison of our values with guideline levels has revealed a polymetallic pollution of sediments in the Ebrié lagoon. Indeed, according to the Ontario Ministry of Environment Screening Guidelines Levels, apart from Mn, concentrations of the other metals might have adverse effects on benthic organisms (Persaud et al., 1990). Effect range severe levels are 250, 10, 75, 110 and 820 mg kg^-1 for Pb, Cd, Ni, Mn and Zn, respectively (Table 5) (Persaud et al., 1990).

Table 3:	Screening quick reference table for heavy metals in marine sediment (Buchman, 1999)

Table 4:	Sediment guidelines and definitions used in SQUIRT (mg kg^-1 dry weight)

Values are in mg kg^-1 dry weight, SQUIRT: Screening quick reference table for heavy metals in marine sediment

Table 5:	Values of the concentrations of heavy metals compared with some sediments guidelines and those determined by other authors (mg kg^-1 dry weight)

NL: Not in literature cited, nd: Not detected. ^aOntario Ministry of Environment Screening Level Guidelines (Persaud et al., 1990), obtain from Praveena et al. (2008). ^bSwedish environmental sediment quality guideline (Mil-Homers et al., 2006)

In sediments collected in the Ebrié lagoon, mean values for Pb, Cd, Ni, Zn and Mn were, respectively 404.75, 2.85, 45.63, 107.46 and 208.77 mg kg^-1 (Table 5). These mean values were ERS range, ERL range, PEL range, PEL range and ERL range, respectively (Table 3-5).

However, according to our data, except for Mn, levels of heavy metals were generally Effects Severe Ranged (ESR) (Table 5) (Persaud et al., 1990; Mil-Homers et al., 2006). Consequently, Pb, Cd, Ni and Zn levels might have adverse effects on animals that live in sediments and waters and could effectively eliminate most of the benthic organisms, due to interactions between sediments and waters (Table 3-5). As example, in the urban area of the Ebrié lagoon, the gastropod Tympanotonus fuscatus fuscatus has disappeared at Bay Biétri, consequence of several hazardous pollutants including heavy metals from anthropogenic inputs (Kone et al., 2008).

Seasonal variations: The seasonal variations of concentrations of heavy metals are presented in Fig. 2. No significant differences (p>0.05) were observed between seasons, apart from Mn. Highest levels for Cd and Ni were observed during the rainy season.

This could be explained by the fact that these elements are introduced into the lagoon through runoff. Manganese concentrations have decreased during flood and rainy seasons. This might result from the input of new sediments or by the fact that hydrodynamism would promote leaching of presents sediments.


Fig. 2:	Seasonal variations of heavy metals levels in the Ebrié Lagoon sediments


Fig. 3:	Variations of Contamination Indexes (CI) and Mean Contamination Index (MCI) of heavy metals

The seasonal variation of zinc concentrations remained low during the study period, suggesting that intakes of zinc might be constants, or dependant on geochemical characteristics of watersheds.

Degree of metal pollution of the different stations: The degree of metal pollution of each site was assessed by the determination of Contamination Index (CI) and the Mean Contamination Index (MCI). The results are presented in Fig. 3.

According to the data, apart from R, located at Bingerville and farther from direct inputs of wastes waters, heavy metals contamination was effective elsewhere as previously observed in waters (Tuo et al., 2012). Highest contamination index for lead were recorded at Unilever, CARENA and Bidet. For cadmium, sediments were most contaminated at Stade FHB, Unilever, Biafra and Marina. Sediments were more contaminated by nickel at Bibet, Unilever and S2. Highest levels of manganese contamination indexes were recorded at S2, Sebroko, Bidet and BNETD. Highest levels of zinc contamination indexes were recorded at Unilever, SIVOA, Grand-caniveau and Stade FHB.

Mean Contamination Indexes (MCI) observed were consistent with the results of the study of spatial contamination with a gradient of pollution close to that of the mold and respecting the following descending rank: Unilever>StadeFHB>Bidet>S2>SIVOA>Sebroko>Grand-caniveau>Marina>Biafra>Abattoir>BNETD>CARENA>Bolibana # SIR>Hôtel-Ivoire>S1>S3>R. This ranking revealed that the waters were most contaminated at areas with high levels of organic matter, domestic and industrial discharges.

Correlations between the parameters were assessed by calculating the Pearson correlation coefficients. The results are presented in Table 6. For physicochemical parameters, positive correlations were observed between pH and temperature (r = 0.64), on one hand and, on the other hand, a high positive correlation between salinity and conductivity (r = 0.90). Indeed, these parameters are characteristics of waters origins in coastal waters such as lagoons. Nickel and zinc were also positively correlated with organic matter. Zinc was positively correlated to Pb, Cd and Ni, suggesting a common source of sediments contamination by these elements (Table 6). Significant positive correlation observed between Ni and Mn might be explained by a common source of inputs of these elements in the Ebrié lagoon or by the nickel adsorption on manganese oxides.

The correlation between the manganese and organic matter was low (r = 0.04), indicating diffuse sources of their inputs in the lagoon. The low affinity between manganese and the organic matter might explain the low concentrations of manganese observed. Indeed, in sediments, organic matter constitutes a trap of choice for many pollutants such as heavy metals. Sequestration of metals by organic matter is through the formation of organometallic complexes relatively stable.

Heterogeneity of studied area: Due to diffuse sources of anthropogenic inputs in the Ebrié lagoon, correlations between sites were assessed with physicochemical characteristics and heavy metals through correlation circle and factorial plan draws using ADE-4 (Fig. 4).

Table 6:	Pearson correlation coefficients between parameters measured in sediments

Temp.: Temperature, Sal.: Salinity, Cond;: Conductivity, O.M: Organic matter, *Significant correlation at p<0.05


Fig. 4:	Correlation circle and Correlation matrix of sediments samples collected in Ebrié Lagoon, O.M: Organic matter, T: Temperature, Sal.: Salinity, Cond.: Conductivity, numbers represent the sites (Table 1)

The axis F1 (horizontal) is generally described by the conductivity, heavy metals such as Zn, Ni and Pb while the axis F2 (vertical) is rather described by the temperature, pH, organic matter, salinity, etc. On axis F1, site 14 (Unilever) and 10 (Bidet) appear to be more affected by heavy metals, unlike sites 15 (S1), 17 (S3) and 18 (R). In fact, sediments at site 14 (Unilever) have recorded the highest concentration of Zn (710.12±36.30 mg kg^-1) while the lowest concentrations of all metals assessed were generally observed at site 18 (R), closed to Bingerville and located in a rural zone. For F2 axis, it can be observed that the maximum value of pH was recorded at site 2 (Sebroko) while the organic matter content was higher in at site 13 (SIVOA) (Table 2). In the Bay Milliardaires, site 16 (S2) was more contaminated by heavy metals compared to sites 15 (S1) and 17 (S3).

CONCLUSION

Heavy metals, Organic matter and some physicochemical parameters were assessed during this study. According to the World Health Organization (WHO, 2004) guidelines, levels of temperature, pH, salinity and conductivity were save for aquatic life. However, sediments have recorded highest levels of organic matter which mineralization might induce anoxia in the ecosystem, therefore unsuitable for all forms of life. According to sediments guidelines, except for manganese and samples collected at Bingerville, sediments collected around Abidjan city were contaminated by heavy metals, with significant differences between sites for Pb, Cd, Ni and Zn (p<0.0001) and seasons for Mn only. Contamination Indexes (CI) and Mean Contamination Indexes (MCI) variations, correlation circle and correlation matrix have revealed the heterogeneity of the studied area, due to the nature of diffuse sources of pollution in Ebrié lagoon. Due to numerous interactions that may exist between sediment and water layers, there is an urgent need to protect the urban area of the Ebrié lagoon. The results of this study can be use as a database for future investigations for the protection of this important ecosystem, found contaminated with heavy metals.