The Sabodala mine site is located in the Kedougou region, in the south-eastern of Senegal, approximately 700 km from the capital Dakar ( Figure 1 ). In this gold complex, ore is extracted from open-pit mines and processed through a conventional crushing-crushing-cyanide leaching cycle at the Sabodala Processing plant. In addition, in April 2024, the mine began commissioning a Bio-oxidation plant (STP / BIOX) to treat refractory ore from one of these large deposits.

The tailings storage facility studied is the first tailings storage facility on the Sabodala site, covering an area of around 388 ha. Following the last dam raise (in 2020), the maximum storage capacity increased to 48.8 Mm³, of which 35 Mm³ was already occupied by tailings in February 2024.

This tailings management facility ( Figure 2 ) is located next to the water basins (Large water dam, small water dam, Upper raw water dam) that the mine uses for workers’ consumption and for ore processing. In this context, it is urgent to monitor the concentration and bioavailability of metals elements in the Sabodala TSF to protect the surrounding water resources.

Tailings samples were collected from the Sabodala tailings facility in February 2024. The samples collected were air-dried and then homogenised by manual mixing. The samples were finely ground (<63 µm) and then digested with a mixture of three acids (perchloric, nitric and hydrofluoric acids). Trace metals were analysed by ICP-AES.

In order to study the affinities of heavy metals with the different soil phases, sequential extractions have been carried out according to the procedure of [4] . The method is based on the use of a set of chemical reagents to release the metals ( Table 1 ). The extraction process was carried out in Erlenmeyer flasks using 1.0 g of the initial dried sample and a precise volume of acid mixture. After stirring, the solution is filtered. The waste is washed with distilled water. The extracts are stored in expensive tubes. In general, the pH of the supernatant from each successive extraction was reduced by 0.5 ml of HNO₃ (63%). Trace metals in extracts are analysed using ICP-AES.

The extractions were carried out in five successive stages (50 ml batch, on 1g of the granulometric fraction smaller than 0.63 mm of a dried and finely ground soil sample). Extracted fractions are as follows:

The order of mobility of ETMs according to their abundance in the different fractions follows the following order: exchangeable > bound to carbonate > bound to Fe-Mn oxide > bound to organic matter or sulphides > residual fraction [4] . Exchangeable elements are a measure of the elements that are most easily released into the environment under acidic conditions and can be considered the most dangerous for the environment [5] . Changes in ionic composition, influencing adsorption-desorption reactions or lowering the pH, could lead to remobilisation [6] [7] , and [8] . Determination of the mobility of the TMEs in the tailings is based on the relative quantity of metal in the exchangeable fraction (T1) containing the mobile forms and in the carbonate-bound fraction (T2) containing the easily mobilised phases. The TME mobility index is calculated by the mobility factor (MF) based on the following equation (1) from [9] [10] , and [11] :

$MF=\frac{T1+T2}{T1+T2+T3+T4+T5}$(1)

As a result, the value of the mobility factor determines the relative mobility and biological availability of the TMEs in the soil.

The average levels of trace elements in mine tailings are 408.1, 79.46, 4.16 and 156.42 ppm respectively for arsenic, antimony, cadmium and nickel. The distribution of these elements in the TSF is presented in Figure 3 . These values are very high compared with the local background (reference points), where the levels are 12.9, 16.64, 0.54 and 92.35 ppm for arsenic, antimony, cadmium and nickel respectively. The data assembled in Table 2 are classified according to the average concentrations of the trace elements studied in mine tailings from five mines around the world. ETM concentrations in our study area exceed UCC standard values [12] .

The data presented in Figure 3 are compared on the basis of their average concentration with the levels of the trace elements studied in the tailings of five mines around the world. The Sabodala mining area, which is the study area, has much lower levels of arsenic, antimony and cadmium than the Tourtit and Ichoumellal mines in central Morocco [13] . The Sabodala mine tailings show greater accumulation than the tailings from mines in Tunisia [14] and Iran [15] . Compared with the Iranian gold mine, the high arsenic and antimony content of the tailings shows that Sabodala's gold is associated with arsenopyrite and stibine ( Table 2 ).

The pollution indices for the tailings and surrounding soils are shown in Table 3 .

The results obtained show that the pollution index values for mine tailings are greater than unity, with the exception of stations S8 and S9. These values indicate polymetallic contamination of mine tailings ( Figure 4 ). The potential risk factor of these metals for the environment (mobility) increases considerably with the combined effect of arsenic, antimony and cadmium. These pollution indices are lower than those found in the Moroccan mines of Mibladen (PI: 34.7) and Zeida (PI: 20.52) [18] .

The Principal Component Analysis (PCA) was applied to gain a better understanding of the geochemical associations characterizing the study area and the possible origins of metal contaminants in mine tailings. After applying the VARIMAX rotation, two main factorial axes were selected for statistical analysis ( Figure 5 ).

Speciation is defined as the determination of a specific form (monoatomic or molecular) or the configuration in which an element can be present in a group of atoms corresponding to different matrices [22] .

Chemical methods are commonly used to determine the speciation of metallic elements in soils. They were used here to complement the physical data obtained and to clarify the potential mobility of metallic elements in the samples studied. The use of chemical extractions is much debated due to certain limitations encountered as a result of (1) refixation phenomena of extracted metallic elements on the solid fraction, (2) the non-selectivity of reagents, (3) the relative efficiency of the extractant depending on the nature of the element and (4) the multitude of protocols proposed. Despite all these limitations, chemical extractions are often used as they provide sufficient information to discuss the relative mobility of metallic elements. In addition, some of the stated limitations can be controlled by returning to the solid residue analysis [23] . To minimize errors, determination using this procedure is carried out three times and then averaged.

Chemical fractionation shows that antimony is preferentially bound to the residual fraction (95%). This close bond was demonstrated by [24] . The proportion of antimony in the sulphide-bound fraction is of the order of 3% ( Figure 6 ). The association of this element with carbonates is relatively low (2%).

Arsenic is found mainly in the residual fraction, with levels of 61%. This is followed by the sulphide fraction (27%), the iron and manganese oxide fraction (7%), the carbonate fractions (4%) and finally the exchangeable fraction (1%). These results confirm the role of iron and manganese oxides in arsenic retention [25] .

Cadmium exists in soils at very low levels, yet it is the most feared element because of its toxicity. Cadmium is mobile, easily bio-available and could either pass into the food chain via plants or migrate deep down to contaminate groundwater. Cadmium is preferentially associated with the residual phase (52%), the sulphide phase with a percentage of around 35% and the carbonate phase with 11%. The exchangeable fractions and those linked to oxides and sulphides are 1% each. This preferential accumulation of cadmium with the residual and sulphide phases was noted in the work of [26] .

Nickel is preferentially associated with the residual fraction ( Figure 6 ), with an average rate of around 51%. This association has been highlighted by several authors [27] [28] .

Bioavailability refers to the capacity of a quantity of an element present in the soil to be absorbed by a living organism [29] . Bioavailability is a tool for assessing toxicity and the risk of pollution. The risk of ecotoxicity by heavy metals does not result solely from the total concentrations found in polluted waste, but also from their mobile fraction [30] [31] . In soils, concentrations of soluble and exchangeable metals are indicators of a potential risk of contaminant transfer to the various environmental compartments [32] [33] . In this study, nickel is the most bioavailable metal, with a 49% share. It was followed by cadmium (48%), arsenic (39%) and antimony (5%) (See Figure 4 ). Metals in form (water-soluble, exchangeable carbonate-bound) are highly mobile and can easily be absorbed and accumulated by living organisms or leached down the soil profile to groundwater [34] , while those embedded in the crystal lattice are relatively immobile and can have little or no harmful impact on living organisms and the environment [35] . Based on the classification described by [36] , if the metal content released into the F1 and F2 fractions of the soil is less than 1% of the total metal, the soil presents no risk to the environment. Values between 1 and 10% reflect a low risk, 11 to 30% a medium risk and 31 to 50% a high risk. Above 50%, the soil presents a very high risk and is considered hazardous, so that TMEs accumulate along the food chain, more in humans than in species lower down the chain [37] . The F1 and F2 fractions of cadmium and nickel in root soil are 12% and 10% respectively, showing a low risk of contamination of groundwater by leaching or of plants by root uptake. Arsenic and antimony show a low risk to the environment.

Equation (1) is used to calculate the mobility factor for metals and metalloids. Figure 4 shows the variations in mobility factors.

Cadmium and nickel can be considered the most mobile elements in mine tailings, as around 12% and 10% respectively are found in the exchangeable and carbonate-bound fractions. This suggests that a substantial proportion of cadmium and nickel is contained in minerals such as calcite and ferrodolomite. The results are in line with those obtained by [38] and [39] . The potential mobility factor for arsenic is 5%. Arsenic is considered to be a rather immobile element because most of this metalloid is bound to the residual mineral fraction. The chemistry of precipitation-dissolution and adsorption-desorption of arsenic is complex and depends on the load of the water, especially when there is a high presence of iron in the water. The potential mobility factor for antimony is 2%. As with arsenic, the distribution of Sb in the tailings indicates that this element was rather immobile and was largely present in the residual fraction. In a recent study on the sequential extraction of antimony from contaminated soils, [24] [40] and [41] found that a large amount (up to 90%) of the antimony was associated with the residual fraction. In general, there are a number of similar general trends in the fractionation of arsenic and antimony in soil. Metalloids (As, Sb) are associated with crystalline and non-crystalline oxide and hydroxide minerals and are often considered immobile [42] .