The study was conducted in the Dakar region, the capital of Senegal and the country’s most densely populated region [8].

Site selection was carried out using the snowball sampling method, which involves tracing connections. Thus, an initial site where WEEE was handled was identified as the starting sample. This group was then asked to identify other WEEE handling sites in turn, and so on. Three occupational categories were considered in this study: collectors, dismantlers, and recyclers.

Soil samples were collected in the towns Mermoz (Mz), Guediawaye (G), Pikine (P), Rebeuss (R), Thiaroye (T), Fass (Fs), and Medina (Ma) (Figure 1).

Figure 1. Geographical location of the sampling sites.

At each identified site, five soil subsamples were collected at a depth of 0 to 5 cm and used to obtain a soil/ash/dust composite representative of the site. Metal wire debris, stone fragments, grass, etc., were removed prior to analysis. The macroscopic appearance of the soils was examined, and the content of trace metal elements (TMEs) was assessed [9]. The samples collected were dried in an oven at 104˚C before being sieved, and samples with a particle size of less than 200 µm were retained for the remainder of the study. A commercially available, uncontaminated soil sample sable de Fontainebleau and a soil sample from a site without WEEE activity (Bambilor soil) were used as controls. During the sampling process, a total of twenty-two composite soil samples were collected. The samples were distributed as follows: two control sites (Bambilor and Fontainebleau), seven collection sites, eight dismantling sites, and five recycling sites.

The pH was determined using a 1 M potassium chloride (KCl) solution at a soil-to-solution ratio of 1:2.5 (w/v). To do this, 5 g of soil was mixed with 12.5 ml of 1 M KCl solution. The mixture was then stirred magnetically for 30 minutes at room temperature. After decanting the resulting mixture, the pH of the supernatant was measured using a pH meter previously calibrated with buffer solutions of pH 4, 7, and 10.

Electrical conductivity, which reflects soil salinity and the overall concentration of dissolved ions, was measured using the distilled water extraction method. For this purpose, a soil-to-water ratio of 1:5 (w/v) was used to ensure consistent reproducibility regardless of the sample’s particle size. The mixture was stirred for one hour, then left to settle. The electrical conductivity of the supernatant was determined using a calibrated conductivity meter. In accordance with the standard, the conductivity results were expressed in millisiemens per centimetre (mS/cm) and denoted as CE₁/₅ to indicate the extraction ratio used. To ensure the accuracy and reproducibility of the data, each pH and electrical conductivity measurement was carried out in triplicate, and the final value corresponds to the average of the three measurements.

The pseudo-total fraction of extractable metals (Cr, Cu, Mn, Fe, Co, Cd, Pb, Zn) was determined using an acid digestion method from the American Public Health Association [10], with modifications to the protocol. Approximately 1 g of soil was transferred to a Teflon tube, followed by the addition of 4 ml of 50% (v/v) nitric acid (HNO₃) and 10 ml of 20% (v/v) hydrochloric acid (HCl). The tubes were covered and placed in a closed-circuit digestion unit on a hotplate for reflux digestion for 30 minutes. After cooling to room temperature, the contents of the tubes were filtered through Whatman No. 41 filter paper. The filtered solution was then transferred to a 50 ml volumetric flask and made up to the mark with distilled water. Solutions were stored at 4˚C until metal analysis by atomic absorption spectrometry.

Analytical quality control:

Metal quantification was performed using a flame atomic absorption spectrometer (Agilent Technologies 240 FS AA). Calibration was carried out with standard solutions prepared from certified stock solutions (1000 µg/mL, Agilent Technologies). Reagent blanks were included in each digestion batch (one blank per 10 samples) and were below the instrumental detection limit for all metals. Precision was assessed by analyzing duplicate samples (n = 3 per metal); relative standard deviations (RSD) were consistently below 10%. In the absence of a certified reference material, accuracy was evaluated through spike recovery tests. Soil samples were spiked with known amounts of each metal prior to digestion, and recoveries were calculated as (measured spiked concentration − measured unspiked concentration) / added concentration × 100. Mean recoveries ranged from 90% to 105% for all metals, confirming satisfactory accuracy. Calibration curves were accepted only if R² > 0.995.

Because the digestion procedure did not include hydrofluoric acid (HF), the measured concentrations correspond to the pseudo-total metal pool, which includes metals from oxides, carbonates, organic matter, and sulfides, but not the residual silicate-bound fraction.

Statistical analyses were performed using R Studio and GraphPad Prism (version 8.00, GraphPad Software, La Jolla, California, USA). Given the non-normal distribution of metal concentrations (assessed by the Shapiro-Wilk test), Spearman’s rank correlation coefficient (ρ) was used to evaluate monotonic associations between metals and soil parameters (pH, electrical conductivity). For each pair of variables, ρ was calculated and a hypothesis test (H₀: ρ = 0) was conducted. Due to the number of comparisons, p-values were adjusted using the Benjamini–Hochberg false discovery rate (FDR) method [11][12]. A significance threshold of q < 0.05 was adopted. The strength of correlations was interpreted based on |ρ|: ~0.1 (weak), ~0.3 (moderate), and ≥0.5 (strong).

Principal component analysis (PCA) was applied to the concentrations of the eight metals (Pb, Cd, Cu, Mn, Fe, Cr, Co, Zn) to explore the multivariate structure of contaminated soils [13]. PCA reduces dimensionality by identifying linear combinations (principal components) that capture the maximum variance. It allowed visualization of similarities between sampling sites and relationships between metals, identification of metal associations indicative of common sources, and distinction between anthropogenic inputs and the natural geochemical background. PCA results were visualized using biplots [14].

The pH and electrical conductivity (EC) measurements (Table 1) highlight variability between the studied sites.

The pH varies from 7.64 (Fontainebleau) to 8.39 (R-G₂), indicating an overall neutral to alkaline environment. The highest values are observed at R-G₂ (8.39) and C-P₂ (8.31), while D-T₂ (7.78) and R-Mz (7.77) exhibit the lowest pH levels. These variations, although limited, can influence the bioavailability of metals, particularly lead and cadmium, whose solubility increases in more acidic conditions.

Table 1.Variation of pH and electrical conductivity.

Note: Each code corresponds to a composite sample from a distinct site (C = collection, D = dismantling, R = recycling, B = Bambilor control, Fb = Fontainebleau control).

Electrical conductivity shows marked heterogeneity, with values ranging from 0.010 mS/cm (Fb) to 4.833 mS/cm (D-P₂). Recycling sites, such as D-P₂ (4.833 mS/cm) and R-Mz (4.433 mS/cm), display the highest levels, suggesting high salinity linked to the accumulation of soluble salts of anthropogenic origin, possibly associated with electronic waste. In contrast, the control sites (Bambilor, Fontainebleau) exhibit low EC values, indicating limited anthropogenic influence.

The analysis of concentrations of Pb, Cd, Cu, Mn, Fe, Cr, Co, and Zn in the soils of collection, dismantling, and recycling sites for WEEE, compared to control sites (Bambilor, Fontainebleau) and international guideline values (WHO, EU, US EPA), highlights variable metal contamination depending on the sites (Figure 2).

The thresholds used for comparison are: for the US EPA, the Regional Screening Levels (RSL) for residential soil [15]; for the EU, the values from Directive 86/278/EEC on the protection of the environment and in particular of the soil when sewage sludge is used in agriculture [16]; for the WHO, although no official soil guidelines exist, the commonly cited values from the WHO/EURO 1996 report on soil quality are used [17]. These international benchmarks were chosen because no national soil quality standards are currently available for Senegal, and they provide a widely recognized reference for assessing contamination levels.

Mn generally shows low concentrations (< 500 ppm), indicating limited anthropogenic influence.

Figure 2. Heavy metal concentrations in soil samples.

The concentrations of Pb far exceed the guideline values at the majority of sites, particularly at D-P₁ (1833.16 ppm), D-G₁ (1339.06 ppm), and D-G₃ (1089.56 ppm), reaching more than four times the US EPA threshold. In contrast, the control sites show very low levels (4.00 ppm in Bambilor and 1.50 ppm in Fontainebleau), confirming the anthropogenic origin of the contamination. Cd generally shows concentrations compliant with WHO and EU thresholds, with some occasional exceedances in certain dismantling areas (D-P₁, D-G₁, R-G₂), but still well below US EPA limits. Cu shows marked enrichments (>1000 ppm) in several dismantling and recycling sites, notably D-G₃, D-P₁, D-P₂, and R-Mz, with exceedances of guideline values. The concentrations of Cr remain below regulatory thresholds, indicating low to moderate contamination, despite some higher levels observed locally. Co presents very low levels, often close to detection limits, indicating the absence of significant accumulation. Zn shows exceedances of WHO/EU values at certain dismantling sites (D-G₁, D-P₂, D-G₃), while remaining below the critical thresholds of the EPA.

Spearman correlations (Figure 3) do not reveal a significant relationship between pH and metal concentrations (|ρ| ≤ 0.27; p > 0.20), probably due to the low variability of pH in a neutral to slightly basic range.

Figure 3. Correlations between physicochemical parameters and metals.

On the other hand, electrical conductivity shows positive correlations, particularly with Cu (ρ ≈ 0.50; p ≈ 0.012) and Fe (ρ ≈ 0.41; p ≈ 0.048), while other metals show weaker relationships (ρ ≈ 0.09 – 0.37). These results suggest that the ionic charge of the medium can influence the mobility of metals through mechanisms of complexation for adsorption sites, and colloidal transport.

PCA (Figure 4) highlights a global gradient of metal contamination carried by the first axis (Dim1), reflecting the intensity of the concentrations. The second axis (Dim2) discriminates between two main groups: (i) Pb-Cd-Zn, associated with anthropogenic sources (batteries, soldering), and (ii) Fe-Mn-Cr-Co, reflecting more of the geochemical background or corrosion processes.

Figure 4. Principal component analysis of metal concentrations at the various sites.

The reference sites exhibit a homogeneous profile dominated by Fe, Mn, and Cr, with low concentrations of anthropogenic metals.

The dismantling sites show marked structuring, with high levels and differentiation between Pb-Cd-Cu-Mn and Cr-Co-Fe-Zn, suggesting chemical signatures related to mechanical and thermal processes.

The collection sites exhibit moderate structuring, indicating diffuse contamination. The second axis opposes Pb-Cd-Zn, linked to electronic components, to Cu-Cr-Fe, associated with alloys or the mineral substrate.

Recycling sites exhibit clear organization, with a separation between Zn-Cd-Mn-Co and Cr-Pb-Cu. This distribution suggests distinct sources, respectively linked to electronic components and thermomechanical activities.

The biplot confirms these trends. Dim1 (57.3% of variance) is dominated by Pb, Cd, Cu, and Zn, indicating a common anthropogenic signature associated with e-waste dismantling. Dim2 (16.4% of variance) is dominated by Fe, Mn, Cr, and Co, reflecting the natural geochemical background. The most contaminated samples project to the positive side of Dim1, while control samples cluster near the origin with low concentrations of anthropogenic metals.