The study was conducted at Jimma University, Jimma, Ethiopia. The wastewater samples and fish were taken from a waste stabilization pond located in the Institute of Technology campus of Jimma University in Jimma town, which is 352 km far from the capital of Addis Ababa in the southwest. The geographic coordinates are approximately 7°30′ latitude north and 36°15′ longitude east. The annual average temperature in the town is 19.3°C (11.5°C–27.1°C), with an annual rainfall of approximately 1,749.1 mm. The WSPs cover an area of 69,236.70 m² (6.9 ha). The facility was designed to cater to a population of over 42,000 people. It consists of seven ponds with a total capacity of receiving more than 2,250 m³/day of raw wastewater. The WSP is designed to serve a population of 40,000 and contains seven ponds, as shown in Figure 1 (21). Initially, the wastewater generated from the student dorms is directed to septic tanks to remove solids, after which it is emptied into WSPs. The untreated domestic wastewater and runoff are combined and disposed of in the waste stabilization pond, resulting in elevated concentrations of various impurities, such as heavy metals, in the influent wastewater. To attain acceptable standard levels, the raw wastewater undergoes both physicochemical and biological treatments to minimize suspended solids and biochemical oxygen demand (BOD).

The waste stabilization pond consists of three components. Each has a distinct size and depth, although the facultative and maturation ponds share the same depth. There are two dimensions to the ponds: anaerobic ponds are deeper and smaller than other ponds. The size and depth are shown as follows:

A = anaerobic pond (2 in number) = 77.94 (L) × 46.49 (W) × 4.75 (D) m

A laboratory-based cross-sectional study design was employed to assess the levels of heavy metals in wastewater and fish tissues (muscle, gill, and liver) collected from the WSPs. Health risk estimation models were employed to assess the human health risks from fish consumption.

The wastewater samples were collected from the surface (0–50 cm), middle (1–1.5 m), and bottom (2.10 m) of the facultative and maturation ponds using a heart valve water sampler, as described in a previous study (22). To ensure the accuracy of the analyses, the samples were then homogenized (23). The wastewater samples were collected in triplicate from each point using 300 ml polyethylene bottles, which were rinsed twice using distilled water, followed by the wastewater samples before the sample collection. The samples were then stored by adding 2.5 ml of concentrated HNO₃ to prevent heavy metal precipitation and algal growth, preserved in the cold chain, and analyzed further in the laboratory of the Department of Environmental Health Science & Technology of Jimma University.

A total of 32 Oreochromis niloticus species were collected from the facultative and maturation ponds. The fish samples were collected by local fishermen using gill nets and wrapped in plastic biosafety bags and then were transported using a cold chain storage to the laboratory for the identification, dissection, and collection of fish tissues (muscle, gill, and liver) for the heavy metal analysis (24, 25).

Wastewater: 100 ml of the filtered wastewater sample was pipetted into a digestion flask with 10 ml of aqua regia (a mixture of HNO₃ and HCl) in a 3:1 ratio. When the sample stopped reacting with HNO₃ and HCl, the flask was placed on a hot plate and heated at 60°C for 30 min. The sample was then allowed to cool before adding 5 ml of H₂O₂ until a clear solution was observed. The content of the flask was transferred into a 50 ml volumetric flask and diluted to the mark of 0.01 N HNO₃ according to a previous study (26)_. All the procedures were performed in the fume hood (27).

Fish: In the laboratory, the samples were cleaned by washing with distilled water. Subsequently, fish tissues, such as muscle, gill, and liver, were separated and sliced into small pieces using a stainless steel knife. Each tissue was then washed with distilled water and dried in an oven at 103°C for 24 h. A 3 g portion was taken into a Teflon beaker and mixed with 5 ml of 70% HNO₃ (Spectrosol). The beaker was placed on a hot plate and heated at 60°C for 30 min. After cooling the beaker, 15 ml of HNO₃ was added and gradually heated to 120°C before removing it from the hot plate once the temperature reached 150°C, turning the contents black. The beaker was allowed to cool before adding 5 ml of hydrogen peroxide (ACS reagent, Aldrich, UK). The sample was then filtered into a 100 ml volumetric flask. The sample blank was prepared in the same manner as the sample preparation. Finally, the treated samples were cooled and filtered using the Whatman No. 42 filter paper. The contents of the beaker were diluted up to 50 ml with 0.01 N HNO₃ for analysis (28–30).

The analysis of the physicochemical parameters in the wastewater samples involved the use of a multiparameter probe (HQ40d) for the field monitoring of the pH, temperature, dissolved oxygen (DO), and electrical conductivity. Turbidity was measured using a turbidity meter (EUTECH TN-100, Singapore). Chloride, total suspended solids (TSS), BOD, nitrate, and phosphate were analyzed in the Department of Environmental Health Science & Technology of Jimma University. The nitrate and phosphate concentrations were determined using a spectrophotometer (DR 5000) at a wavelength of 410 nm and 690 nm, respectively (31).

Heavy metal analysis: Heavy metals (Pb, Cd, As, and Hg) were detected in the pre-treated samples of wastewater and fish tissues through micro plasma atomic emission spectrometry (Agilent 4210 MP-AES) for Pb and Cd and hydride-generated atomic absorption spectrometry (HGAAS, novAA 400P, Germany) for As and Hg. The prepared sample was sent to the Ethiopian Food and Drug Authority and the Ethiopian Construction Design and Supervision Work Authority in Addis Ababa for laboratory analysis. The wavelength and detection limits of each heavy metal were as follows: 405.781 nm and 0.005 mg L⁻¹ for Pb; 193.7 nm, 228.80 nm, and 0.005 mg L⁻¹ for Cd; 253.7 nm, 253.65 nm, and 0.0001 mg L⁻¹ for Hg; and 193.7 nm and 0.0001 mg L⁻¹ for As.

Bioaccumulation factors: The accumulation of toxic metals in fish was calculated with a bioaccumulation factor (BAF) as the ratio of the concentration of an individual heavy metal in the fish tissues to the concentration of the metals in the water (32). The BAF was estimated based on a study done by Kumar (16).(1)BAF=ConcentrationofHMindrytissuesoffish(μg/L)ConcentrationofHMinpondwater(μg/L)

The possible public health risks of toxic metals through fish consumption from the study area were evaluated. The data and reference values for the input parameters and the heavy metal contents detected in the fish muscles used for the human health risk assessment are summarized in Table 1.

The estimated daily intake (EDI) was used to assess the susceptibility to heavy metals in children and adults from direct consumption exposure pathways. The EDI of heavy metals from muscle consumption was assessed by using the concentration of heavy metals in fish samples, daily ingestion of fish by the community, and average Body weight (BW) by using the method followed in the previous studies (39, 41):(2)EDI=FIR∗MC∗CFWABFIR is the daily fish consumption rate. Although people around fish sources are usually meat consumers, consumption lifestyles have currently been changing to fish in regions and societies where fixed and adequate fish stock is available. In such societies, the annual fish ingestion can surpass 10 kg/capita (19). Accordingly, the daily mean fish consumption rates (CR) projected for adults and children were considered to be 0.03 and 0.016 kg/person/day, respectively (33). These data were taken from previous studies because of the absence of national per capita fish consumption databases in the country.

MC is the concentration of heavy metals (by milligrams per kilogram) detected in fish meat.

BW_av denotes the mean body weights set by the WHO as 15 kg and 60 kg for Ethiopian children and adults, respectively (34).

In terms of carcinogenic risk, the following model predicted the maximum permitted daily ingestion rate or limit (CR_lim) of metals in fish (kg day⁻¹) (40):(3)CRlim=ARL∗BWCSF∗MCwhere CR_lim is the maximum allowable daily consumption rate or limit of contaminated fish (kg day⁻¹); BW is the mean BW of the consumer population (kg); RfD stands for the oral reference dose (mg kg⁻¹day⁻¹); and MC is the metal concentration in the edible part of fish (mg kg⁻¹) (39).

In terms of non-carcinogenic risk, the maximum allowed daily intake of fish was estimated using the following equation (40):(4)CRlim=RfD∗BWMCwhere CR_lim is the most permissible daily consumption rate or restriction of contaminated fish tissue (kg day⁻¹); BW is the mean BW of the consumer population (kg); RfD is the oral reference dose (mg kg⁻¹ day⁻¹); and MC is the concentration of heavy metals determined in fish tissue (mg kg⁻¹).

The non-carcinogenic risk was explored using the target hazard quotient (THQ), which is a measure of the risk level (non-carcinogenic) associated with pollutant exposure. The hazard quotient was computed using the following equation (40):(5)THQ=EF×ED×FIR×CF×MCRfD×WAB×ATnThe oral reference dose (RfD) for Cd = 0.001, Pb = 0.004, As = 0.0003, and Hg = 0.0001 (37).

While the oral RfD = oral reference dose of a substance (mg/kg/day) based on the oral intake for every metal for a grown-up client with an average BW of 60 kg, THQ < 1.0 points out that negative fitness chances are not likely to occur. If the THQ is greater than or equal to at least one, it is most probable that an adverse health hazard could happen (42).

Exposure to two or more metal contaminants may have additive health consequences. Therefore, the cumulative health risk was calculated by summing the THQ [also known as the hazard index (HI)] (43).(6)HI=(THQ(Pb)+THQ(Cd)+THQ(As)+THQ(Hg))A value of HI greater than 1 shows a greater health risk.

The target cancer risk (TR) was computed to assess the carcinogenic effects. The method for estimating the TR is also available in the USEPA Region III Risk-Based Concentration Table (37). The model for estimating the TR is as follows (29):(7)TR=MC×FIR×CSF×ED×EFBW×ATnwhere CSF is the cancer slope factor (mg kg⁻¹ day⁻¹), which was 1.5 mg kg⁻¹ day⁻¹ for As, 0.0085 mg kg⁻¹ day⁻¹ for Pb, and 0.38 for Cd, while the other parameters have been defined previously. The TCR was estimated for As, Cd, and Pb since these elements may promote both non-carcinogenic and carcinogenic effects depending on the exposure dose. As and Cd are known as Group A carcinogens, and Pb is known as a Group B carcinogen.

The collected data were coded and entered into Epi Data version 3.1 before being exported to SPSS (version 26) for analysis. Descriptive analysis was conducted for each variable, and correlation analysis was performed to analyze the intermetallic association and the significance level with physicochemical parameters. A paired t-test was conducted to determine the statistical significance between two sampling points. To evaluate whether heavy metal concentrations in fish tissues varied based on the site, a one-way analysis of variance was employed at a significance level of α = 0.05. The concentrations of all metals were reported in micrograms per liter for wastewater and in milligrams per kilogram on a dry weight basis for fish tissues.

To maintain the quality of the data, laboratory instruments were calibrated, bank measurements were performed, and triplicate analyses were made for each sample. Stock standard solutions (1,000 mg L⁻¹) containing 2% HNO₃ of Pb, Cd, Hg, and As (Buck Scientific Puro-Graphic) were used. The calibration curve was determined using serial dilutions. All the reagents used in the experiments were analytically pure. The samples were then digested without delay and analyzed by HGAAS and MP-AES following documented procedures.

The study was carried out following acceptance of an ethical clearance letter from Jimma University's Institute of Health Science. To get the necessary cooperation for the study, the Institute of Health, Jimma University, issued a letter of support to the Institute of Technology, Jimma University and laboratory facilities.

The assessments of various physicochemical parameters, namely, pH, temperature, turbidity, BOD, DO, conductivity, TSS, chloride, nitrate, and phosphate, were conducted using standard methods as described in the American Public Health Association (44). The average pH and DO values increased from the inlet to the outlet of the pond, while the other parameters showed a decreasing order along the wastewater treatment process (Table 2).

The calibration curves for Cd, Pb, As, and Hg were obtained by using standard solutions prepared from their respective stock solutions. The correlation coefficients of heavy metals ranged from 0.996 to 0.999, indicating a strong relationship. The correlation coefficients of the heavy metals were determined using prepared standards vs. their corresponding absorbance. Finally, quality results were obtained from the sample analysis for each heavy metal using HGAAS and MP-AES.

The heavy metals were found to be in decreasing order of concentration (µg L⁻¹), that is, Cd > Pb > As > Hg, in both the facultative and maturation ponds. Cd showed the highest concentration among the heavy metals throughout, while Hg had the lowest concentration. The maximum levels of all heavy metals were obtained from the facultative pond. Furthermore, the Cd and Pb concentrations exceeded the acceptable levels established by the WHO (32) and the USEPA (36), which may pose a huge threat to public health and the natural environment (Table 3).

The correlation analyses performed on the data enabled the identification of the possible common characteristics of heavy metals in wastewater. Cd exhibited a negative correlation with As (r² = −0.314, p < 0.05) and a positive correlation with Hg (r = 0.239, p < 0.05). Pb showed a strong positive correlation with Hg (r² = 0.779) and a moderately positive correlation with As and cadmium (r² = 0.229, 0.415). The significant positive correlations obtained among some heavy metals proved that they may have similar accumulation behaviors or originate from the same environmental sources of pollution (45).

The association between heavy metals and other physicochemical parameters has also been assessed using the Pearson correlation coefficient (r), and the association was statistically significant at two-tailed (p < 0.01). Pb showed a strong positive correlation with the pH, EC, turbidity, BOD, Cl, PO₄⁻³, and NO₃, and the association was statistically significant at two-tailed (p < 0.05). Moreover, Hg and Pb showed a strong negative correlation with DO. The significant positive correlations of Hg and Pb with other physicochemical parameters may confirm that a considerable share of wastewater parameters is associated with the adsorption and oxidation of trace metals. The lack of a significant correlation between heavy metals and other wastewater parameters might be caused by the compositional variety controlling heavy metals. A paired sample t-test analysis revealed no significant variation for the heavy metals being studied, except for Hg, which demonstrated a substantial statistical difference between facultative and maturation ponds (t = 6.450, p = 0.023).

Biometric data, such as the length and weight of O. niloticus fish collected from the WSPs, were measured. The average length and weight of O. niloticus (n = 32) collected from the facultative and maturation ponds were 24.76 ± 0.64 cm and 108.33 ± 7.89 g, 16.54 ± 1.68 cm, and 82.33 ± 4.67 g, respectively. It can be used to provide information concerning the aquatic health and environment.

The heavy metal concentrations expressed as mg kg⁻¹ dry weight were detected in the muscle, liver, and gill of O. niloticus through HGAAS and MP-AES. In the liver, O. niloticus accumulated the highest concentration of Pb, which ranged from 0.34 mg kg⁻¹ to 0.37 mg kg⁻¹, and the gill exhibited the highest value of Cd, which ranged from 0.345 mg kg⁻¹ to 0.406 mg kg⁻ in the maturation and facultative ponds, respectively. However, the muscle showed the lowest concentration of all heavy metals. It can be noticed that different organs exhibited different patterns of accumulation, that is, gill, Cd > Pb > Hg > As; liver, Pb > Cd > Hg > As; and muscle, Pb > Cd > Hg > As from both ponds (Figures 2 and 3).

The results of the analysis of variance revealed that there are significant differences in the metal concentrations in the different fish tissues (p < 0.05). Mercury was selectively accumulated in all types of fish tissues, with a statistically significant difference at p < 0.05. The Cd concentration showed significant differences between the liver and the muscle (p < 0.05), whereas As and Pb did not show significant differences between the fish tissues. It indicates the accumulation efficiency of any particular metal in fish tissues.

The highest concentrations of lead and cadmium were recorded in the muscle of O. niloticus. The arsenic concentration in the muscles was above the MPL of 0.01 mg kg⁻¹ as the FAO/WHO recommended standard (46) (Figure 4).

The bioaccumulation of toxic metals in the fish tissues collected from the WSPs was determined. The maximum BAF of Hg in the liver of O. niloticus was calculated using Equation 1, which was 1,302.5 and 764.73 at the maturation and facultative ponds, respectively. It was observed that the bioaccumulation factors were increased in the order of Cd < Pb < As << Hg for all tissues (Table 4). This is a clear indication that the bioaccumulation factors of Hg in all O. niloticus tissues were higher compared with other metals.