The study area (Figure 1) is located in the region of ‘‘Haut Sassandra’’ (Centre: West of Côte d’Ivoire) between latitudes 6˚20 and 5˚45 North and longitudes 7˚15 and 6˚30 West. The climate is of the subequatorial type (hot and humid) characterized by four seasons including the long rainy season with off-seasons and storms (April to mid-July), the short dry season (mid-July to mid-September), the short rainy season (mid-September to November) and the long dry season (December to March). However, the duration of the dry seasons clearly gains ground on that of the rainy seasons because of the destruction of the forests by anarchic and abusive exploitation. The average rainfall which fluctuated between 1500 and 1700 mm is only 1200 mm currently. The average annual temperature is 25˚C (Die, 2006).

The study was conducted in six districts of the city of Daloa and a village located about 25 km from the city. A total of ten (10) water points including nine (9) drinking water points and one river water were sampled. The drinking water sampled is of four types: traditional well water, spring water, borehole water and tap water. These different types of water are the main sources of drinking water in the study area. As for the Lobo River, it is used as raw water for the production of drinking water from the taps. The sampling sites and their characteristics are given in Table 1. A total of 11 water quality parameters were analyzed. Water samples were taken in 1.5 liter polyethylene bottles for physico-chemical parameters and 0.5 liter for heavy metals. Each bottle is rinsed with the water to

*DMS (Degrees Minutes Seconds).

be collected then filled to be stored in a cooler containing ice. However, a few drops of concentrated hydrochloric acid were added to the 0.5 L samples intended for the analysis of heavy metals.

The pH was determined in situ using a Horiba LAQUAtwin PH-11 portable pH meter. Magnesium (Mg²⁺) and calcium (Ca²⁺) ions were determined by complexometric assay using EDTA with NET (Black Eriochrome T) and Patton and Reeder as indicators. Total hardness (TH) in water samples was determined using Equation (1) in which Ca²⁺ and Mg²⁺ concentrations were expressed in mg/L (Zakir et al., 2020). The fluoride (F⁻) content in the water samples was determined by potentiometry using a pH/Ionometer XL250 as described by Yapo et al. (2021). The heavy metals copper (Cu), zinc (Zn), cadmium (Cd), iron (Fe), manganese (Mn) and lead (Pb) were determined using an atomic adsorption spectrometer in flame mode air-acetylene (Varian SpectrAA 20).

HT = ( 2.5 × Ca 2 + ) + ( 4.1 × Mg 2 + ) (1)

Metal loading (ML) is defined as the arithmetic sum of the concentration of all metals assayed in the sample. The ML (mg/L) for each water sample was calculated with the mathematical formula given by Equation (2) (Zakir et al., 2020):

ML = ∑ i = 1 n C m = C Cd + C Cu + C Fe + C Mn + C Pb + C Zn (2)

The Metal Pollution Index (MPI) is suggested as a reliable and accurate method for monitoring metal pollution (Liu et al., 2019). It makes it possible to compare the total heavy metal content between the different study sites (Sanou et al., 2021b; Sanou et al., 2022). The MPI was calculated with the equation proposed by Usero et al. (1997):

MPI = ( C 1 × C 2 × ⋯ × C n ) 1 / n (3)

where C represents the concentration of the metal in the sample and n the total number of metals assayed in the water sample.

The Contamination Index (CI) is generally used to assess the cumulative effect of metal content in waters against recommended international standards for drinking water (Backman et al., 1998; Rajmohan, 2022). In the present study, CI is used to explore fluoride and metal contamination. The CI is calculated using Equation (4) (Rajmohan, 2022):

CI = C A i C N i − 1 (4)

The Comprehensive Pollution Index (CPI) is a comprehensive pollution index method applied to assess qualitatively overall water quality. CPI has been used to classify water quality status by several authors (Zhao et al., 2012; Mishra et al., 2015; Kaioua et al., 2022). The global pollution index is evaluated by Equation (5) (Zhao et al., 2012):

CPI = 1 n ∑ i = 1 n C i S i (5)

where C_i is the concentration of contaminant i, S_i is the standard norm of contaminant i, n is the number of parameters and CPI is the global pollution index. The permissible standard concentrations of each parameter considered in the study were obtained from WHO standards (WHO, 2017). The CPI varies from 0 to 2 and is used to classify the level of water quality as shown in Table 2.

Calculation models have been recommended by the United States Environmental Protection Agency (USEPA) to assess the hazards to humans of various contaminants in air, food and water. Among the contaminant exposure pathways, oral ingestion of contaminants such as drinking water consumption is a major concern (Rajmohan, 2022). In the present study, the health risk through the consumption of water contaminated with fluoride and heavy metals was calculated for adults and children.

To establish a relationship between the physico-chemical parameters, fluoride ions and heavy metals, the Bravais-Pearson correlation test was performed using STATISTICA software (Version 7.1). The linear correlation coefficient r of Bravais-Pearson varies from −1 to +1. The value −1 indicates a perfect negative correlation and the value +1 represents a perfect positive correlation while the value 0 shows an absence of correlation between the parameters (Kouyaté et al., 2021).

Table 5 presents the values of the physico-chemical parameters of the drinking water sampled. Drinking water in the study area is acidic in nature with pH values ranging from 4.4 to 6.53. Only tap water has a pH within the pH range (6.5 - 8.5) recommended for drinking water (WHO, 2017). Raw water from the Lobo River is alkaline in nature (7.18) and falls within the allowable range for water for human consumption. Non-compliant values observed at groundwater levels could lead to irritation of eyes, skin and mucous membranes (Kalankesh et al., 2022). The concentrations of calcium and magnesium in drinking water vary from 2.40 to 28.05 mg/L and from 0.07 to 0.68 mg/L respectively. These values are below the maximum limit authorized by the WHO in drinking water, which is 75 mg/L for Ca²⁺ and 50 mg/L for Mg²⁺ (Salam et al., 2021). The total hardness recorded values that oscillate between 3.29 mg/L and 74.72 mg/L. According to Udhayakumar et al. (2016), the maximum value of total hardness allowed for consumption purposes is 600 mg/L. However, hardness above 300 mg/L can lead

to heart and kidney problems (Udhayakumar et al., 2016). With regard to the results of total hardness, all the water samples were identified as being suitable for human consumption. All tap water samples (100%) comply with the potability standard, indicating that the raw water treatment technique used by the Ivorian water distribution company is adequate for these parameters.

The fluoride concentrations in the sampled waters oscillate between 0 and 2.4 mg/L (Figure 2). The P1, S1 and F1 sites do not contain fluoride (0 mg/L). According to He et al. (2020), a low concentration of fluoride (F < 0.5 mg/L) in drinking water causes dental caries in infants and children, as well as osteoporosis.

The P2, S2, F2, R2 and R3 sites have contents which fluctuate between 0.8 and 1.3 mg/L. These concentrations remain below the standard (<1.5 mg/L) suggested by the WHO (WHO, 2017) and could be beneficial for consumers. Indeed, fluoride concentrations between 0.4 and 1.0 mg/L in drinking water has beneficial effects on teeth, especially in young children, as it promotes calcification of dental enamel and protects teeth against dental caries (Loganathan et al., 2013; Yapo et al., 2021). The R1 tap is not suitable for consumption. Indeed, this water has a fluoride content (2.4 mg/L) higher than the standard suggested by the WHO. Consumption of water with a high fluoride concentration (F > 1.5 mg/L) for a long period can lead to dental fluorosis (Adewole et al., 2021). Concentrations above the standard have already been encountered in borehole water at Boguédia in the department of Daloa (2.20 ± 0.01 mg/L) by Yapo et al. (2021).

The heavy metal contents in the drinking water samples taken varied from 2.56 to 4.47 for Cd, from 3.14 to 4.74 for Cu, from 4.93 to 10.70 for Fe, from 2.74 to 5.91 for Mn, from 4.42 to 11.66 for Pb and from 1.93 to 20.91 for Zn. Taking into account the standard norms (Table 6) recommended for drinking water by

the WHO, the results recorded show that all the water samples are inappropriate for consumption except the well P2 and the borehole F1 in the case of Zn. These high concentrations can create adverse health effects such as cancer, hypertension, lung disease, gastrointestinal bleeding, kidney disease, neurological disorders and reproductive effects (Bhan & Sarkar, 2005; Zakir et al., 2020). The results indicate that the treatment of the distribution company effectively reduces the content of the metals Cd, Cu and Pb. However, there is an increase in the content of Fe, Mn and Zn in all the tap waters in comparison with that of the raw water (Table 7). This presence of metals at a high rate is probably due to the aging of the equipment of the supply or connection network pipes on the one hand (Douard & Lebental, 2013) and on the other hand to the inefficiency of the treatment. Metal loading (ML) reflects the total amount of metals assayed in a sample. The metal load is higher in tap R3 (50.19 mg/L) and lower in borehole F1 (26.09 mg/L). The order of accumulation of metal load in drinking water is: R3 > R1 > S1 > R2 > P1 > S2 > F2 > P2 > F1 (Table 7). The Metal Pollution Index (MPI) was used to compare the overall content of heavy metals in the sampled waters. The values recorded indicate overall contamination in the following decreasing order: R3 > S1 > R2 > R1 > P1 > S2 > F2 > P2 > F1 (Table 7).

However, these values are greater than 1 (MPI > 1); thus reflecting that all the metals contained in drinking water are not without danger for human consumption (Abdel Ghani, 2015; Sanou et al., 2021b). In general, the treatment provided by the water distribution company reduces overall metal pollution. Indeed, the MPI of the Lobo River, source of raw water for the production of drinking water, is higher than the MPI of all the taps (Table 7).

In the waters sampled, the values of the IC vary from −1 to 0.6 with an average of −0.47 ± 0.51 for F⁻; from 852.33 to 1489 with an average of 1180 ± 212.82 for Cd; from 0.57 to 1.37 with an average of 0.97 ± 0.27 for Cu; from 15.43 to 34.67 with an average of 26.72 ± 6.74 for Fe; from 5.85 to 13.78 with an average of 10.16 ± 2.52 for Mn; from 441 to 1165 with an average of 719.3 ± 247.10 for Pb; from −0.36 to 5.97 with an average of −2.67 ± 2.19 for Zn. According to Rajmohan (2022), IC value ≤ 0 indicates drinking while water samples with IC value > 0 indicate unpleasant drinking water as shown in Table 8.

The CI for fluoride is less than zero in 88.88% of samples whose consumption is advised. Regarding heavy metals, all CIs are greater than zero in the samples studied. The consumption of these waters is therefore not recommended. However, 37.5% of the samples have IC < 0 for Zn which makes them acceptable as drinking water. The complete pollution index (CPI) has values ranging between 241.89 and 443.11 with an average of 324.39 ± 63.97. These values are well above 2 (CPI > 2); indicating severe pollution of the waters studied (Mishra et al., 2015) which makes them unfit for human consumption. The high CPI values could be attributed to the high concentration of iron, lead and zinc recorded during the study.

The Bravais-Pearson correlation matrix was created to elucidate the relationships between the variables measured in the different water samples. The correlation coefficients are presented in Table 9. The coefficients in bold are significant at p < 0.05. Examination of the matrix indicates the existence of positive and significant correlations between the physico-chemical parameters, fluorides

*The coefficients in bold are significant correlation at p < 0.05.

and heavy metals: pH - Ca²⁺ (r = 0.63), pH - Mg²⁺ (r = 0.82), pH - TH (r = 0.67), Ca²⁺ - Mg²⁺ (r = 0.52), Ca²⁺ - TH (r = 0.99), Ca²⁺ - F⁻ (r = 0.70), Ca²⁺ - Fe ( r = 0.53), Ca²⁺ - Zn (r = 0.59), Mg²⁺ - TH (r = 0.58), TH - F⁻ (r = 0.68), TH - Fe (r = 0.53 ), TH - Zn (r = 0.59), Fe - Zn (r = 0.82). These observed significantly positive correlations between the variables indicate the mutual dependence of the parameters on each other, a common source or similar behavior during transport (Shetaia et al., 2020; Sanou et al., 2021a, 2021b). Furthermore, significant negative correlations were found between Ca²⁺ - Cd (r = −0.62), TH - Cd (r = −0.59), F⁻ - Cd (r = −0.58), Cd - Fe (r = −0.77), Cd - Zn (r = −0.82), Cu - Fe (r = −0.77), Cu - Zn (r = −0.82). These results show an inverse dependence between these variables, thus reflecting the decrease in one of the parameters with the increase in the other (Sanou et al., 2020; Kouyaté et al., 2021; Sanou et al., 2021b). Non-significant correlations (r < 0.5) were observed between certain variables. These low correlations indicate that the presence or absence of one of these parameters has little effect on the content of the other (Kam et al., 2019).