Speciation of trace metal elements in sediments of Alibori River, North-West of Benin

Alibori River located in northwestern Benin is subject to contamination by trace metal elements (TME: Zn, Cu, Pb, Cd) due to the use of agricultural inputs in its vicinity. This study focuses on the speciation of TME elements in the sediments of this river. The sediments were sampled during the rainy and dry seasons, dried, crushed and then sieved. The sediments were then mineralized and subjected to sequential fractionation. The total extraction and sequential extractions were analyzed using a spectrophotometer UV. The sediments of Alibori River had an acid pH and were contaminated by TME during the dry season. The zinc and copper had an affinity with the different fractions with a high proportion in the fractions F1 and F2 during the rainy season and a high proportion in the F5 fraction during the dry season. The lead and cadmium had an affinity with the F5 fraction during the rainy season, and the F3 and F4 fractions during the dry season. The calculated contamination index, individual contamination factor and risk assessment code (RAC) demonstrated that the river presented a potential risk of water contamination. The aforementioned parameters also showed that the river induced risk of bioavailability ranging from medium to very high for organisms. The TMEs found in the sediments of the Alibori River were linked to the exchangeable and acid-soluble fractions which made them more mobile and bioavailable to organisms.


INTRODUCTION
Rivers are natural moderately important water course, flowing continuously or intermittently following a defined route towards an ocean, sea, lake or another river. The Alibori River is an ecosystem located in the Sudanian *Corresponding author. E-mail: chitounafiou@gmail.com. Tel: 00229 96798646.
Author(s) agree that this article remain permanently open access under the terms of the Creative Commons Attribution License 4.0 International License zone between 10° 30 "and 12° North latitude and 1° 32" and 3° 50 'East longitude in North Benin. It extends in the department of Alibori with many tributaries scattered in the municipalities of Kandi, Gogounou, Banikoara, Ségbana, Karimama and Malanville before flowing into the Niger River (Le Barbe et al., 1993;Thierry et al., 2016).
Alibori River's waters are used by the populations for daily needs (drink, cooking, laundry, dishes) and the processing of agricultural products as adequate local water supply facilities are not available in the region (PEA) (Tachégnon and Ibrahim, 2012;Chitou et al., 2019). Intensive agriculture is also developed around the river and is dominated by cash and food crops such as cotton, and corn (Agagbé, 2008;Thierry et al., 2016;Tachégnon and Ibrahim, 2012). It is characterized by an intensive use of agricultural inputs (chemical fertilizers) and phytosanitary products.
One of the consequences of the use of these products (chemical fertilizers and phytosanitary products) is the contamination of waterways by various pollutants such as pesticides and trace metal elements (TME). These TMEs can lead to morphological abnormalities and neurophysiological disorders, and affect enzymatic and hormonal activities in living organisms (Chitou et al., 2019;Wang et al., 2017).
There is a dearth of information on the effects of human activities on the Alibori River (Thierry et al., 2016;Chitou et al., 2019). Our previous study showed that sediments of the Alibori River were contaminated by TMEs (Zn, Cu, Pb, Cd) due to the use of agricultural inputs in its surroundings (Chitou et al., 2019). This study also determined the total concentration of TMEs in the river's sediments.
The total quantification of TMEs available in sediments remains insufficient to predict (or evaluate) their toxicity, reactivity, mobility and bioavailability (Zhou, 2009;Daldoul et al., 2015;Bashar, 2016). These parameters are rather evaluated through the chemical speciation of ETMs. The current study is therefore initiated to assess the toxicity, reactivity, mobility and bioavailability of TME in sediments of the Alibori River. Its objective is to determine the different chemical fractions of Zn, Cu, Pb, and Cd available in the sediments of this river during each season.

Collection of sediment samples
Seven (7) sediment-sampling points were chosen in Alibori River based on the position of the crop fields and various human activities in its vicinity (Figure 1), and the uses of its waters (Table 1). The river's sediments were sampled during each of the two seasons (rainy and dry) prevailing yearly in the Alibori region. The sampling Chitou et al. 143 of sediments was performed using a Schipeck grab or manually with a gloved hand depending on the water level.

Sediment analysis
The seven (7) sediments were dried to a constant weight using an oven at 105°C, crushed, while preserving the shells, and then sieved at 2 mm. The obtained product was stored at 18°C for analysis. The pH of the processed sediments was measured as described by standard AFNOR NF X-31-103, 1998. The sediments were then mineralized by adding 20 ml of a mixture of hydrochloric acid (HCl) 6 mol.L -1 and nitric acid HNO3 7 molL -1 (3: 1) to 1 g of each sample. The mineralized sediments were brought to the boil on a hot plate for 1 h at 120°C, cooled, then centrifuged, and the resulting supernatant was recovered. The collected supernatant was combined with the appropriate volume of distilled water to obtain a solution of 25 ml, and then analyzed using a spectrophotometer UV to determine the total content of TME.

Sequential extractions of TME
The speciation of TME was carried out according to the sequential method of Tessier et al. (1979) modified for 1 g of sediment. The sediments were thus subjected to fractionation according to the sequential extraction as described in Table 2. The extracted fractions were then analyzed using a spectrophotometer UV.
Given the spatial distribution of the seven (7) sites, and the contamination index CI of these sites, only the sediments from the sites 2, 3, 5 and 6 underwent fractionation (speciation).

Data processing
The levels of variations of the physicochemical parameters and the concentrations of copper, zinc, cadmium and lead in the sediments, as well as extractions' results, were collapsed into histograms using the Excel software.
The Risk Assessment Code (RAC) and the Individual Contamination Factors (ICF) were estimated and used to assess environmental risks and possible impacts on the river's benthic organisms.
The heavy metal data were subjected to principal component analysis (PCA) and correlation calculation using X-Lstat software. The correlation coefficient was used to determine the relationship between two variables and to measure the strength of this link. The calculation of the correlation coefficient was used for screening of the possible relations between metals with regard to common origin, homogeneous distribution, identical behavior concerning physicochemical processes, and similar reciprocal fractionation.
The contamination index "CI" was estimated following Equation 1 and used to assess the degree of pollution of the sediments in line with Table 3  . (1) where TMEs: concentration of TME in the Alibori River sediments,  mg.kg -1 ; TMEn: concentration of TME considered as normal, mg.kg -1 .
The risk of water contamination by a pollutant was examined using the individual contamination factor (ICF) of sediments by TME, which was estimated following Equation 2. For . ≤ ICF≤ 69.42, there are potential risks of water contamination by TME (Lawani, 2013). ICF = Concentration of non-residual TME fractions / concentration of residual TME fraction (2) The concentration of non-residual fractions is the sum of the fractions F1, F2, F3, and F4. Fraction F5 is the residual fraction.
The availability of metals in solution was assessed using the classification of the Risk Assessment Code (RAC) outlined in Table  4 (Huajun et al., 2011;Abdallah, 2012). The RAC is based on the percentage scale to the exchangeable and carbonate fractions of TME in sediments and is the sum of the exchangeable and carbonate (acid-soluble) fractions.

Zinc and copper
The zinc and copper contents varied in the sediments at each site according to the seasons (Figures 2 and 3). The concentrations of both metals were higher in the sediments during the rainy season than the dry season except the site 6 for zinc and sites 3 and 4 for copper. The high zinc content was observed at site 7 during the rainy season and site 6 during the dry season.
The highest copper's content was observed at site 7 during the rainy season and site 3 during the dry season.
The zinc and copper contents varied in the sediments of each site according to the seasons (Figures 2 and 3). The concentrations of both metals were higher in the sediments during the rainy season than the dry season except the sites and for copper. The highest zinc's content was observed at site 7 during both seasons. The highest zinc's content was observed at site 7 during the rainy season and site 3 during the dry season.

Lead and cadmium
Lead's concentration was higher in the sediments during the rainy season than the dry season at the sites 1, 2 and 3 ( Figure 4). From sites 4 to 7, the reverse phenomenon was observed.
Cadmium was almost absent in the sediments during the rainy season except the site 7 which had a high content (52.5 mg.kg -1 sediment) ( Figure  5). During the dry season, cadmium's concentrations in the sediments were low.

Contamination index (CI)
The Zn and Cu contamination index (CI) was well above 3 (Table 8). The CI of Cd and Pb was less than 3 except for cadmium during the dry season and at site 7 during the rainy season.

Fractions of zinc
Zinc was distributed in all geochemical fractions of sediments during the rainy season. The residual fraction was less observed at the first three sites (Bassi ''S2'', Kpéwari ''S3'', Deroubou ''S5''). The residual fraction was the most observed (47%) at site 6 (New bridge). At site 3, zinc was much more bound to the exchangeable and acid-soluble fractions which were the most mobile sensitive to a small change in pH. At site 5, the oxidizable fraction was dominant. At site 2, the oxidizable and reducible fractions were the most observed ( Figure 6).
Zinc was predominantly present in the residual fraction during the dry season except the site 3  (Huajun et al., 2011, Abdallah, 2012 Criteria related to sediment (%) Risks <1 No risk 1-10 Low risk 11-30 Medium risk 31-50 High risk >50 Very high risk    (Kpéwari) where zinc was more related to the reducible fraction ( Figure 7). Figure 8 shows the distribution of copper in the different geochemical fractions of the Alibori River's sediments during the rainy season. Except for the residual fraction, the copper was much more bound to the reducible and acid-soluble fraction at the sites 2 (Bassi) and 3 (Kpéwari). At site 5 (Dèroubou), all the fractions were moderately observed. At S6 site (New Bridge), the copper was much bound to the oxidizable and reducible     fractions. During the dry season, the residual fraction was the most observed with a percentage of 68% (site 2) ( Figure  9). At site 3, zinc was weakly bound to the exchangeable and residual fraction. Site 5 was dominated by the oxidizable and exchangeable fractions. At site 6, the share of the acid-soluble fraction was the highest (53%). Figure 10 shows the distribution of lead in all geochemical fractions during the rainy season. A low proportion of the residual fraction was observed at the sites 2, 5 and 6 and a high proportion (51%) at site 3. The lead was much more bound to the oxidizable fraction (46%) at site 6. At site 2, lead was bound to all fractions, unlike other sites.

Fractions of lead
During the dry season, an increase in the oxidizable fraction was observed up to 52% (sites 2 and 3) for the lead (Figure 11). The lead also exhibited affinity with the reducible and acid-soluble fraction during this period at the level of the sites 5 and 6.

Fractions of cadmium
From Figure 12, cadmium was much bound to the residual fraction. Besides, it was linked to the  exchangeable fraction, especially at the sites 3 and 6, and to the acid-soluble fraction at site 2.
Cadmium was strongly bound to the reducible and oxidizable fractions at all sites and the acid-soluble fraction at the sites 2, 5 and 6 ( Figure 13). The exchangeable fraction was weakly observed.

Individual contamination factor (ICF) of TMEs
The calculated contamination factor varied from 4.96 to 11.16 for zinc, 1.82 to 57.47 for copper, 0.96 to 7.76 for lead and 0.3 to 1.66 for cadmium. The highest ICF value was observed at site 5 for zinc, site 3 for copper and site 6 for lead and cadmium during the rainy season. During the dry season, the calculated contamination factor was between 1.2 to 4.36 for zinc, 0.47 to 15.31 for copper, 1.63 to 8.99 for lead and 2.02 to 15.1 for cadmium. The highest value of the ICF of copper and zinc was observed at site 3.

Risk assessment code (RAC)
The calculated RAC for zinc and copper were in the range % <RAC≤6 % with high risk at sites and during the rainy season and the site 6 during the dry season. The calculated RAC for lead and cadmium were  in the range of % <RAC≤ 9. %. The highest RAC was observed at the sites 5 for Pb and 3 for Cd during the rainy season and the site 6 during the dry season for Cd and Pb. Figure 14 shows the results of the PCA during the rainy season. The majority of the parameters were well observed because they were close to the correlation circle with a proportion of 81.14% including 43.48% for the F1 axis and 37.66 for the F2 axis. It was observed a strong correlation between the oxidizable fraction of zinc (Znspox) and organic matter (OM), an average correlation between the oxidizable fraction of copper (Cuspox) and organic matter. It was also observed a strong correlation between the exchangeable fraction of zinc (Znspech) with the F1 axis and with RAC of zinc (ZnspRAC), likewise, there was a correlation between the acid-soluble fraction of copper (Cuspacdb) and RAC of copper (Tables 9 and  10).  It is well known that the square that accompanying the correlation circle reflects various dominant variables responsible for the pollution at each site. This suggests that the site of the new bridge was dominated by the oxidizable fraction of cadmium and lead. The Kpéwari site was dominated by the exchangeable fraction of cadmium and the acid-soluble fraction of copper. The Bassi-Alibori site was very close to the centre therefore it was dominated by the exchangeable fractions of zinc, acid-soluble zinc and copper, and the oxidizable fraction of copper. The Dèroubou site was dominated by the acidsoluble fraction of lead and the oxidizable fraction of zinc. Figure 15 outlines the results of the PCA during the dry season. The majority of the parameters were well observed during the dry season because they were close to the correlation circle with a proportion of 78.32% including 47.20% for the F1 axis and 31.12 for the F2 axis. There was a strong correlation between the oxidizable fraction of cadmium and the individual contamination factor of copper and zinc as well as the F1 axis. These parameters and the axis were correlated with the oxidizable and reducible fractions of zinc (Znssox, Znssrdt). There was a strong correlation between the acid-soluble fraction of copper (Cussacdb), the acidsoluble fraction of lead (Pbssacdb) and RAC of lead. The Bassi Alibori site was dominated by the residual fraction of coppers (Cussresd), zinc (Znssresd) and cadmium (Cdssresd), and the reducible fraction of cadmium (Cdssrdt).

Principal component analysis (PCA)
The Dèroubou site was dominated by the exchangeable fraction of zinc (Znssech) and the reducible fraction of zinc. The Kpéwari site was characterized by the exchangeable fractions of lead, oxidizable from cadmium and zinc, reducible from copper and zinc. The site of the new bridge was characterized by exchangeable fractions of cadmium, acid-soluble of lead and copper.

DISCUSSION
Chemical fertilizers and pesticides used in agriculture release TME into the environment, which can enter aquatic ecosystems through runoff (Nziguheba and Smolders, 2008;Yamaguchi et al., 2009;Alassane et al., 2018). Depending on the physicochemical conditions of the environment, these TME integrate suspended matter and are deposited overtime on the sediments which constitute pollutants sources.
The values recorded for the pH of the Alibori River's sediments in this study suggest that the environment was slightly acidic during the rainy season and more acidic during the dry season, and governed the behavior of TME between the solid and liquid phases. The increase in acidity was caused by the decomposition of organic matter of which concentrations have decreased from 44.98 to 13.50 mg.kg -1 , producing CO 2 . The observed pH values were lower than those reported for sediments at the coastal lagoon of Benin (6.9 to 8.7) by Waris et al. (2017), and at the level of the Boumerzoug Algeria basin (6.15 to 8.51) by Leila et al. (2014). The divergence between our results and those of these authors could be explained by several factors including the difference between the sampling periods and the activities developed around the bodies or courses of water. Zinc, copper and lead were more concentrated in the sediments during the rainy season, which is a cultivation period, than the dry season, unlike cadmium. This observation is in agreement with Chitou et al. (2019) who reported similar variation for TMEs concentrations in Alibori River's sediments between the dry season and the rainy season in 2018.In accordance with Chitou et al. (2019), it could be concluded that the high values of TME during the cultivation period were due to the use of fertilizers and pesticides since the river is located in a rural area dominated by agricultural activities (Chitou et al., 2019). The concentration of Zinc and copper in the sediments has increased compared to that obtained in the 2018 during the dry season (Max: 15.6 mg.kg-1 for zinc, 19.1 for copper) and the rainy season (Max: 175 mg.kg -1 for zinc, 231 mg.kg -1 for copper) using a metalyser (Chitou et al., 2019). This increase could be due to the accumulation of TME in the sediments and the analysis method. This accumulation depended on the proportion of non-labile and strongly retained fractions in the sediments. The observed cadmium during the dry season may be explained by the degradation of organic pollutants (pesticides) which can last for weeks depending on environmental conditions. Therefore, agricultural activity through the use of chemical inputs and pesticides remains the main source of contamination of the Alibori River's sediments by Cu, Zn, Cd, and Pb, without however excluding domestic discharges and atmospheric deposits from the combustion of fuels associated with road traffic.
The estimated contamination indices (CI) confirmed the contamination of the sediments by copper, zinc, lead and cadmium in the order Zn˃ Cu˃ Cd˃ Pb. The CI values were overall greater than 10, this suggests that the Alibori River constituted a high-risk area for living beings (abiotic factors) according to the sediment contamination assessment grid (Aouatif et al., 2016).
This contamination index remains insufficient to assess the risks of contamination of aquatic species and fauna by TME in the environment. Rather, these risks were assessed through the chemical speciation of ETMs.
The speciation analysis of the TME showed that zinc and copper were present in all fractions with varying proportions F ˃ F ˃ F ˃ F ˃ F during the rainy season. By ignoring site 6 (new bridge) because of its position in cotton fields and other crops, the order of the fractions changes F ˃ F ˃ F ˃ F ˃ F for zinc and F ˃ F ˃ F ˃ F ˃ F for copper which explains the effect of the proximity of the culture fields to these sites.
The TME linked to these different fractions (F1, F2, F3, F4) of the sediments can be released into the water depending on the physicochemical conditions of the environment to be bioavailable. The bioavailability involves more the metal species linked to the F1 and F2 fraction because they are labile and can be toxic to aquatic organisms Daldoul et al., 2015. The F1 and F2 fractions had the highest proportions suggesting that copper and zinc were the most labile, thus bioavailable for aquatic organisms. This observation shows that the contamination of the sediments by zinc and copper constituted a risk for the aquatic fauna and flora of Alibori River. The affinity of copper with the F2 fraction is in agreement with previous study on the copper in sediments of the Ouémé delta (Rodrigue et al., 2018). The observed affinity confirms the effect of pH on the behavior of TMEs in the aquatic environment. The acid-soluble (F2) and exchangeable (F1) fractions predominated at low pH (5 -5.86), and the reducible (F3) and oxidizable (F4) fractions appeared when the pH increased. This is in line with Bashar et al. (2016) who reported that the precipitation of cations is favorable at high pH. The retention of TME in the different fractions of the sediments as indicated in reaction Scheme 1 may explain the accumulation of TME in the sediments during the rainy season.
During the dry season, the reverse reaction was observed F ˃ F ˃ F ˃ F ˃ F for zinc and F ˃ F ˃ F ˃ F ˃ F for copper, which could be explained by the release of the bound zinc and copper to these fractions into the liquid matrix (water). This chemical reaction which took place between the solid (sediment) and liquid (water) matrix was due to the variation in pH.
Zinc adsorbed on the particles (F1) and the part bound to carbonate (F2) were sensitive to a slight decrease in pH. The decrease in pH observed during the dry season making the environment acidic lead to a decrease in these fractions (F1, F2) at certain sites, which may justify the low zinc content in the sediments at these sites.
Indeed, according to Serpaud et al. (1994), the absorption of Cu and Zn decreases when the temperature increases (30 -40°C). However, the pH being inversely proportional to the temperature, therefore the decrease in pH suggest to the increase in temperature of the medium, thus causing the decrease in the absorption of copper and zinc. There was likely a phenomenon of chemical release as shown in reaction Scheme 2, which explains the increase in the concentrations of the latter (TME) observed in the water in previous work in this period (Chitou et al., 2019). There is a great affinity between organic matter and metallic trace elements (Zhao et al., 2012), which explain the strong correlation observed between the oxidizable fraction of copper, zinc and organic matter during the rainy season as shown by the circle of correlation. The proportion of the oxidizable fraction (F4) observed during the dry season was due to the gradual degradation of organic pollutants.
Lead and cadmium were also unequally distributed in all fractions as follows: F ˃ F ˃ F ˃ F ˃ F for lead F ˃ F ˃ F ˃ F ˃ F for cadmium, with a much greater enrichment in the residual fraction during the rainy season, unlike copper and zinc. During the dry season, the oxidizable and reducible fraction is the most dominant following the order F ˃ F ˃ F ˃ F ˃ F for lead F ˃ F ˃ F ˃ F ˃ F . The observed affinity of cadmium and lead proves that they were less mobile, unlike copper and zinc. The distribution of cadmium and lead in the different fractions was weakly influenced by pH. Metals (Cd and Pb) are more bioavailable at low salinity (Zhao et al., 2012). The salinity of the waters of Alibori River is zero, which explains the strong affinity of cadmium and lead with the F3 and F4 fractions and hence their weak bioavailability.
The individual contamination factors (ICF) of Zn, Cu, Cd and Pb recorded during this study were in the range (0.14; 69.42), which proves that the contamination of sediments by TME constituted potential risks of contamination for the river's water. The calculated RAC (Risk Assessment Code) showed a risk of bioavailability ranging from medium to very high for these TMEs. The bioavailability of these TMEs for the aquatic environment is linked to their affinity with the exchangeable and acidsoluble fractions as the binding (particles-metals) and (carbonate-metals) is weak in these fractions (Perin et al., 1985).
This affirmation supports the strong correlation between Znspech and ZnspRAC, Cuspacd and CuspRAC, Pbssacd and PdssRAC, Cdssech and CdssRAC as revealed by the circles of correlations. The problem of bioavailability risk posed by these TMEs for the aquatic environment is believed to cause the disappearance of taxa from the Alibori River as reported by Thierry et al. (2016). Indeed, according to Chouti (2011), these benthic species do not tolerate an average concentration of 0.77 mg/kg cadmium. The risk of bioavailability can lead to the disappearance of aquatic species owing to the anomalies caused by the TME in their body (Okoro et al., 2012;Wang et al., 2017).

Conclusion
The sediments of the Alibori River were contaminated with metallic trace elements. The results of our analyzes is revealed an accumulation of these metallic trace elements in the sediments with concentrations greater than the standard for Zn, Cu and Cd and high contamination indices.
Therefore, the chemical quality of the sediments constituted a risk for the aquatic resources of the Alibori River.
The study of the speciation of Zn, Cu, Cd and Pb showed that these elements were distributed in all fractions of sampled sediments with a high proportion in the exchangeable fraction and acid-soluble for zinc and copper, particularly during the rainy season. The fractions of Zn and Cu were however quite mobile, suggesting that these elements may be readily available to organisms. In addition to metals, pesticides also contain active molecules which are persistent toxic pollutants and very dangerous for aquatic species and humans. Thus, it will be relevant to screen the contamination of the Alibori River by active molecules from pesticides under the next stage of our work on the assessment of the impact of the use of agricultural inputs on this river.