1 Introduction

Fish are regarded as inexpensive protein sources for communities residing in the vicinity of rivers and dams (Esilaba et al., 2020). However, fish consumption is reported to be one of the routes of metal exposure to humans due to their capacity to accumulate metals when inhabiting contaminated environments (Kumar et al., 2022; Pal & Maiti, 2018). The accumulation of metals may be influenced by numerous factors i.e. pH, conductivity, temperature, salinity, fish age and size, sex, and the presence of other metals (Bawuro et al., 2018). According to Jezierska and Witeska (2006), fish length may exhibit an inverse relationship with metals due to dilution whereas the metabolic capacity of smaller fish makes them susceptible to metal accumulation. Moreover, different fish organs accumulate metals differently depending on their affinity, location relative to the medium, and the duration of exposure. The liver, kidney, gills, and spleen have shown to accumulate higher concentrations compared to muscle (Debipersadh et al., 2023; Gilbert et al., 2017; Njinga et al., 2023). However, muscle is included in metal accumulation studies as it is the part consumed by humans (Şirin et al., 2024).

Metals such as Sb, cadmium (Cd), Cr, molybdenum (Mo), Ni, and Pb are potential threats to human health and they are also documented by the United States Environmental Protection Agency (USEPA) as metals of concern during the Anthropocene era (Heath et al., 2004; US-EPA, 2000). Studies reported increased concentrations of these metals in water bodies impacted by agricultural activities (Han & Gu, 2023), wastewater effluents (Soleimani et al., 2023), mining (Lakra et al., 2019; Mishra et al., 2008), acid mine drainage (Lebepe et al., 2020), and industrial activities (Adegbola et al., 2021). It is, therefore, imperative to assess the behaviour and concentrations of these metals in aquatic biota, particularly those receiving water from impacted catchments.

The uMgeni River system is one of the important freshwater systems supporting over 10 million people in South Africa. Its middle stretch which includes the Nagle and Inanda dams is experiencing extensive artisanal fishery by local communities. Studies reported high metal concentrations of Sb, Cd, Cr, Mo, Ni, and Pb in the surface water and sediment in Inanda Dam which were associated with the inflow from a highly polluted uMsunduzi River (Mdluli et al., 2023; Olaniran et al., 2014). These metals were found to be problematic in most river systems draining industrialized catchment (Adegbola et al., 2021; Hao et al., 2022), and they are among the metals that resulted in fish advisories in other parts of the world (Heath et al., 2004). However, no study was conducted in the uMgeni River to evaluate the safety upon consuming the inhabitant fish. Therefore, the present study aims to evaluate concentrations of these selected metals in one of the preferred fish species, Oreochromis mossambicus (Peters, 1852). and check whether they are safe for human consumption. The first hypothesis was that fish from both dams would exhibit a common descending trend of liver > gill > muscle for all metals. Based on the location of the Inanda and Nagle dams relative to the uMsunduzi River, we also hypothesized that metal concentrations would be higher in fish from Inanda Dam compared to Nagle Dam with fish from the former exhibiting concentrations exceeding acceptable levels for human consumption.

2 Materials and Methods

2.1 Sampling Localities

Fish were sampled at Inanda and Nagle dams of the uMgeni River system in KwaZulu-Natal, South Africa. The Inanda Dam is bordered by rural communities that practice subsistence fishery to supplement their protein needs. Moreover, the dam receives water from a heavily polluted uMsunduzi River and the two smaller streams (Fig. 1) (a: uMngcweni River and b: Sikelekehleni River). The water quality in the Inanda Dam has been deteriorating over the past few decades (Gumbi et al., 2017). Olaniran et al. (2014) recorded elevated metal concentrations in Inanda Dam. In contrast, Nagle Dam is located approximately 49 km upstream, before the uMsunduzi-uMgeni rivers confluence. Moreover, the dam is known for its better water quality compared to Inanda Dam due to negligible anthropogenic activities along the river (Adeyinka, 2014; Oosthuizen & Ehrlich, 2001).

Fig. 1
figure 1

The uMgeni River catchment map with dams of interest encircled

Full size image

2.2 Water and Sediment Quality Assessment

The physical parameters of water were measured using the HANNA multi-parameter instrument during wet and dry seasons in 2020 and 2021. Water samples were then collected using polyethylene sampling bottles and refrigerated until processing and analysis. During analysis, water was acidified using hydrochloric acid and filtered through a 0.45 µm membrane. Filtered water was analysed for metals using Inductively coupled plasma mass spectrometry (ICP-MS, Agilent 7900). The guidelines by the Canadian Council of Ministers of the Environment (CCME, 2001, 2012) and the Department of Water Affairs and Forestry (DWAF, 1996) were used for evaluating water quality.

Similarly, sediment samples were collected using AMS Shallow Water Bottom Dredge sampler. Three grabs were made at each site and mixed to form a composite and stored in a 1 L polyethylene sampling bottles. Three composites were collected from each site. Samples were kept in the fridge for metal analyses. Sediment processing was done following the protocol described in Misra et al. (2024). Approximately 0.2 g of sediment samples were digested using aqua regia, 3-hydrochloric:1-nitric acid (3HCL:1HNO3) supplemented with 30% hydrogen peroxide (H2O2) for digestion of organic matter. The digested solution was double filtered with 0.45 µm pored membranes and syringe filters. The filtered sample was topped to 250 ml mark using deionized water. Metal analysis was done using ICP-MS (Agilent 7900).

2.3 Fish Sample Collection

Fish were caught using gillnets with different mesh sizes (45 to 90 mm stretch) and electro-shocker. Fifteen fish were caught at Inanda Dam over two seasons whereas 12 were caught at Nagle Dam. Fish were euthanized by cutting through the spinal cord and weighed using a digital spring scale. The lengths were measured using a measuring tape. Fish were then opened ventrally for harvesting of other organs such as liver and gills. Muscle tissue was also harvested as it is the primary organ consumed by humans.

2.4 Fish Sample Digestion and Metal Analysis

Approximately 0.2 g of fish tissues were sectioned and digested following the method described in Misra et al. (2024) using aqua-regia. Samples were heated until the brown smoke faded and then allowed to cool at room temperature. After cooling, the samples were double filtered using 0.45 μm pore membranes and syringe filters, then topped off to the 250 ml mark with deionized water. Solutions were kept in the fridge until analysis using ICP-MS.

2.5 Quality Control

All samples were digested with blanks and multi-element standards were also used during analyses. The standards ranged from 0.01 – 5 mg L−1 for Sb, Cd, and Pb, and 0.1 – 10 mg L−1 for Cr, Mo, and Ni. The accuracy of the analytical technique was determined by analysing certified reference materials (CRMs) with every 5 samples analysed. For fish tissues, the certified reference materials for fish (DORM-4) were used. The recovery in all these analyses ranged from 93—103% (Table 1) and the detection limit of the instrument was 0.001 mg L−1.

Table 1 The linear concentration, the limit of detection, the limit of quantification, and the recovery percentage observed during analysis
Full size table

2.6 Bioaccumulation Factor

The bioaccumulation factor (BAF) is the ratio of metal concentration in the organ to the medium, either water or sediment (Arnot & Gobas, 2006). It may be used to determine whether bioaccumulation has occurred (Melake et al., 2023). It was calculated using the Eq. 1 as per Arnot and Gobas (2006).

$$\text{BAF}=\frac{\text{Concentration in organ }(\text{mg}/\text{kg})}{\text{Concentration in medium }[\text{sediment }(\text{mg}/\text{kg})\text{ or water }(\text{mg}/\text{l})]}$$
(1)

2.7 Non-Carcinogenic and Carcinogenic Risk Assessment

The non-carcinogenic risk assessment was carried out by calculating the target hazard quotient (THQ) following US-EPA (2000) protocols as described in Gilbert et al. (2017) and Lebepe et al. (2020). Parameters used for calculating the THQs are presented in Table 2. The calculation of THQs were done using Eq. 2 whereas the carcinogenic risk index (CRI) was calculated out using Eq. 3.

$$\text{THQ}=\frac{\text{C x EF x ED x IRF}}{\text{RfD x BW x AT}}$$
(2)
$$\text{CRI}=\frac{\text{C x EF x ED x IRF x OSF}}{\text{RfD x BW x AT}}$$
(3)

where C is the metal concentration in mg/kg wet weight, EF: exposure frequency (days per year), ED: exposure duration (years), IRF: ingestion rate (kg per week), RfD: reference dose (mg/kg), BW: body weight (kg), AT: average time (days) and OSF: oral slope factor. The non-carcinogenic health risks are probable if the THQ is > 1 whereas no health risk are expected for THQ of < 1 (US-EPA, 2000). The slope factor available in the literature is only for Cd, Cr, and Pb with the maximum acceptable index being 10–4 (US-EPA, 2013).

Table 2 Parameters used for calculating the target hazard quotient (Gilbert et al., 2017; Heath et al., 2004; Lebepe et al., 2016)
Full size table

2.8 Data Analysis

Statistical analysis was done using various packages of R-4.3.1 (R Development Core Team, 2014). Homogeneity of variance and normality were determined using Levene’s and Shapiro–Wilk tests, respectively. Depending on the satisfaction of the assumption, the Mann–Whitney U Test or independent sample t-test was used to compare the difference between means of concentrations in sediment and fish tissues. The non-metric multidimensional scaling (NMDS) was performed using a vegan package (Oksanen et al., 2013). The correlation was determined using Pearson or Spearman’s tests using the PerformanceAnalytics package (Peterson et al., 2020). Statistical results were considered significant at p < 0.05.

3 Results

3.1 Metals in the Water and Sediment

Antimony and Cd showed mean concentrations of 1.48 mg L−1 and 0.001 mg L−1 in the water column at Inanda Dam whereas other metals were below detection level at both dams. However, sediment exhibited notable metal concentrations at both dams (Fig. 2). Antimony was significantly higher at Inanda Dam compared to Nagle Dam (w = 20, p < 0.05). Similarly, Cd exhibited a significantly higher concentration at Inanda Dam compared to Nagle Dam (w = 25, p < 0.05). No significant differences were observed for Cr (w = 25.5, p > 0.05) and Ni (w = 6, p > 0.05) between the two dams. In contrast, a significantly higher concentration was observed for Mo at Inanda Dam (w = 25, p < 0.05). Coinciding Ni trend, Pb has also showed a significantly higher concentration at Inanda Dam (w = 31, p > 0.05).

Fig. 2
figure 2

Box and whisker plot for metal concentrations observed for sediment at Inanda and Nagle dams

Full size image

4 Metal Concentrations in Fish Tissues

4.1 Comparison of Metals in Organs Between Dams

Metal concentrations observed in the tissues of O. mossambicus are shown in Fig. 3. The gill showed significant differences for Sb (t = 2.31, p < 0.05), Cr (t = 3.32, p < 0.05), Mo (t = 3.87, p < 0.05), Ni (t = 4.59, p < 0.05) and Pb (t = 2.35, p < 0.05) with Sb and Pb exhibiting relatively higher concentrations at Inanda Dam and Cr, Mo and Ni showing higher concentrations at Nagle Dam (Fig. 3). In contrast, no significant difference was observed for Cd (t = 1.52, p > 0.05) in the gills between the two populations. In the muscle, significant differences were observed for Sb (t = 2.34, p < 0.05), Cr (t = 2.32, p < 0.05), Mo (t = 2.39, p < 0.05), Ni (t = 1.95, p < 0.05) and Pb (t = 2.97, p < 0.05). A trend similar to that of the gills was observed for muscle where Sb and Pb showed relatively higher concentrations in the Inanda Dam population with Cr, Mo and Ni showing higher concentrations at Nagle Dam. Moreover, Cd concentration in the muscle showed no significant difference between the two populations (t = 1.66, p > 0.05). A different trend was observed for the liver with only Cr (t = 2.28, p < 0.05) and Mo (t = 2.57, p < 0.05) showing significant differences between the two populations. Higher concentrations for both metals were observed in the Inanda Dam population compared to Nagle Dam. No significant differences were observed for Sb (t = 1.77, p > 0.05), Cd (t = 1.74, p > 0.05), Ni (t = 0.72, p > 0.05) and Pb (t = 0.16, p > 0.05) between the two dams.

Fig. 3
figure 3

Box and whisker plot for metal concentrations observed in three tissues of Oreochromis mossambicus from Inanda and Nagle dams

Full size image

Coinciding with the trends described, NMDS showed no clear separation for metal concentrations in the liver (p > 0.05) between the two dams and the dispersion was also not significant (p > 0.05) with average distances to median being 0.35 for Inanda Dam and 0.37 for Nagle Dam. In contrast, gills and muscle exhibited clear separations between the two populations (p < 0.05) with non-significant dispersion within each group (Fig. 4). The gill average distance to median was 0.29 for Inanda Dam and 0.24 for Nagle Dam whereas the average distance to median in the muscle was 0.25 in Inanda Dam and 0.28 at Nagle Dam.

Fig. 4
figure 4

Non-metric multi-dimensional scaling depicting metal concentration in organs between dams (∆ Inanda; ▲ Nagle) (upper row) and each dam between organs (▲ liver; ∆ gill, ○ muscle) (lower row)

Full size image

4.2 Comparison of Metals Between Tissues in Each Dam

Descending trends liver > gill > muscle for Sb, Ni and Pb, muscle > liver > gill for Cd, liver > muscle > gill for Cr and muscle > gill > liver for Mo were observed at Inanda Dam (Fig. 3). At Nagle Dam, descending trends liver > gill > muscle for Sb, Cd, Ni and Pb, and muscle > gill > liver for Cr and Mo were observed at Nagle Dam (Fig. 3). An NMDS showed liver being separated from the gill and muscle at Inanda Dam (p < 0.05) (Fig. 4) with average distances to median being 0.35, 0.29 and 0.31 for the liver, gills and muscle, respectively (p > 0.005). In contrast, dispersion at Nagle Dam showed a significant difference between organs (p < 0.05) (Fig. 4) with average distances to median being 0.37, 0.22 and 0.44 for the liver, gill and muscle, respectively. Nevertheless, MANOVA showed a significant difference between groups (p < 0.05) (Fig. 4).

4.3 Bioaccumulation Factor

Bioaccumulation factors were not calculated for tissue-water concentrations as metals were below detection levels in the water. Sediment showed notable concentrations, hence, bioaccumulation factors were calculated. The bioaccumulation factor for tissue-sediment concentrations were generally higher at Nagle Dam compared to Inanda Dam (Table 3). Only Cr in the liver and muscle exhibited a higher bioaccumulation factor at Inanda Dam (Table 3).

Table 3 The bioaccumulation factor for biota-sediment (BAFs) observed for selected tissues of Oreochromis mossambicus from Inanda and Nagle dams
Full size table

4.4 Inter-Metal and Metal-Fish Length Correlation

The gills and muscle showed strong to moderate significant relationships for Cr–Mo, Cr-Ni, Mo-Ni and Sb-Pb whereas fish weight showed significant moderate relationships with Cr, Mo and Ni (Figs. 5 & 6). In the liver, a moderate significant relationship was observed for fish weight and Sb (Fig. 7).

Fig. 5
figure 5

Correlation coefficients for metal concentrations in the gill of Oreochromis mossambicus

Full size image
Fig. 6
figure 6

Correlation coefficients for metal concentrations in the muscle of Oreochromis mossambicus

Full size image
Fig. 7
figure 7

Correlation coefficients for metal concentrations in the liver of Oreochromis mossambicus

Full size image

4.5 Non-Carcinogenic and Carcinogenic Risk Assessment

The non-carcinogenic THQs observed for O. mossambicus from Inanda and Nagle dams are presented in Fig. 8. Over 50% of the population at Inanda Dam and over 25% of the population at Nagle Dam exhibited THQ > 1 for Sb. Chromium is also THQ > 1 for approximately 25% of the Inanda Dam population and over 75% of the Nagle Dam population. Molybdenum was within the threshold value of 1, however, approximately 10% of the Nagle Dam population was > 0.5 (Fig. 8). The Ni THQ was less than 0.5 at both dams whereas Pb showed THQs exceeding the threshold value of 1 at both dams. Over 75% of the population at Inanda Dam exhibited THQs > 1 whereas just over 25% of Nagle Dam population exhibited THQs > 1 (Fig. 8). The CRIs were generally higher at Inanda Dam compared to Nagle Dam. The Inanda Dam populations exhibited CRIs of 1.67 × 10–1, 3.13 × 10–2, 2.45 × 10–4 for Cd, Cr and Pb, respectively whereas the Nagle Dam population showed CRIs of 3.53 × 10–2, 6.36 × 10–2 and 1.10 × 10–4 for Cd, Cr and Pb, respectively.

Fig. 8
figure 8

Target hazard quotients observed for Oreochromis mossambicus at the Inanda and Nagle dams

Full size image

5 Discussion

5.1 Metals in the Water and Sediment

The uMgeni River is known for its poor water quality, particularly in the lower stretch that receives water from a tributary that drains a highly urbanized and industrialized catchment, the uMsunduzi River (Misra et al., 2024; Olaniran et al., 2014). In the present study, the neutral to alkaline pH was observed at both dams. The observed pH was complemented by metals showing undetectable concentrations in the water column and significant concentrations in sediment. The pH plays a crucial role in metal solubility and mobility in aquatic ecosystems (Arnous & Hassan, 2015). At non-acidic pH, metals precipitate and sink to the bottom sediment (Saalidong et al., 2022), which could be the explanation for the observed metal trend between the water column and bottom sediment at both dams. The Inanda Dam exhibited relatively higher metal concentrations in sediment compared to the Nagle Dam. The Inanda Dam is located downstream of the uMgeni-uMsunduzi rivers confluence (Misra et al., 2024). Given that the uMsunduzi Rivers drains a highly industrialized catchment, it is evident that its impact on the water quality is reflected in the Inanda Dam. This trend of metal concentrations between the two dams is comparable to that reported by Mdluli et al. (2023) in these dams. Moreover, these metal concentrations were comparable to those observed by Egbe et al. (2023) in wastewater-impacted River Meme in Cameroon. Moreover, concentrations reported in the sediment at both dams are substantially lower than those reported by Sojka and Jaskuła (2022) in numerous rivers impacted by industrial and agricultural activities in Poland and those reported by Lebepe et al. (2020) in an acid mine drainage impacted Olifants River in South Africa. Nevertheless, the Cd concentration observed in the present study was higher than that reported by Algül and Beyhan (2020) in wastewater and agriculture-impacted Lake Bafa in Türkiye. However, all metal concentrations in sediment were within the CCME (2001) guidelines recommended for freshwater ecosystems. Nevertheless, the anthropogenic activities impacting metal concentrations in the Inanda Dam are a cause for concern given that the dam is used for subsistence fishery by local communities.

5.2 Metal Concentrations in Fish Tissues Between Dams

Fish can accumulate metals with various organs being target sites for metal taken up through different routes (Njinga et al., 2023). Gills are in direct contact with water and were found to be a target site for waterborne metals (Zaghloul et al., 2024). Despite the water column showing concentrations below the detection limit for most metals, the gills of the Inanda Dam population showed relatively higher concentrations for Sb and Pb with Cr, Mo, and Ni being higher at Nagle Dam. Lead and Cr concentration ranges observed at both dams were lower than those observed by Pan et al. (2022) in numerous fish species from the metal-polluted Yellow River in China with Sb, Cd, and Ni being comparable. In contrast, Cd was lower than those reported by Zaghloul et al. (2024) in numerous fish species with Pb being comparable. Moreover, Cd, Cr, Ni, and Pb concentrations in the gills of the Inanda Dam population were comparable to those observed by Jayaprakash et al. (2015) in rivers impacted by wastewater works and industrial activities and exhibiting significantly low metal concentrations in the water column and substantial concentrations in sediment. The Cd concentrations were lower, with Cr and Pb being higher than those reported by Njinga et al. (2023) in the gills of Tilapia brevimanus from waters exhibiting insignificant concentrations of these metals. Gills are susceptible to waterborne metals and are among the components that sequestrate metals from the water column in polluted aquatic ecosystems, which could be the explanation for higher concentrations in the gill even when metals are significantly low in the water column.

The liver is a primary organ for detoxification, hence, a target site for metals that have already entered the body (Pan et al., 2022). This was also evident in the present study with most metals showing elevated concentration at both dams. Chromium and Mo exhibited higher concentrations in Inanda Dam than in Nagle Dam whereas other metals showed no difference between the two populations. Despite the liver being a site for detoxification, its potential to reflect concentration in the water body is poor as it does not respond rapidly to metal increase in the outside environment but stores accumulated metals (Rajeshkumar & Li, 2018). These dynamics of metal accumulation in the liver could explain the non-significant difference observed in the liver for most metals between the two dams. Nevertheless, the observed concentrations were lower than those observed in rivers impacted by wastewater works and industrial activities (Arantes et al., 2016; Blankson et al., 2024; Jayaprakash et al., 2015) and comparable to those observed by Zaghloul et al. (2024) in a non-polluted Bardawil Lake in Egypt. In contrast, Cr, Cd and Ni were higher than those reported by Kotacho et al. (2024) in Lates niloticus and Oreochromis niloticus from a waterbody impacted by industrial and agricultural activities. Moreover, Cd and Pb concentrations were relatively higher than those reported by Hasan et al. (2023) in numerous fish species from wastewater effluents-impacted Turag River in Bangladesh, with Cr being lower. Metal concentrations accumulating in the liver showed no definite trend in relation to concentrations in the water or sediment (Hasan et al., 2023; Jayaprakash et al., 2015; Yang et al., 2023). The metal concentration in the liver did not reflect the trend observed in sediment between the two dams. This could be expected as metals reaching the liver could be through both dietary and external environmental exposure.

Despite the muscle location relative to the external water environment, it does not respond rapidly to metal increase in the water column as it does not play any metabolic role (Blankson et al., 2024; Pan et al., 2022). However, a trend similar to that observed for the gill was observed in the muscle with Sb and Pb showing relatively higher concentrations in the Inanda Dam population and Cr, Mo, and Ni showing higher concentrations at Nagle Dam. Metal concentration ranges observed in the present study are comparable to those observed in other related studies on Labeo umbratus (Gilbert et al., 2017), Labeo rosae, and O. mossambicus (Jooste et al., 2015; Lebepe et al., 2020), Tilapia brevimanus and Euthynnus alletteratus (Njinga et al., 2023), and Oreochromis niloticus and Coptodon kottae (Egbe et al., 2023) from metal polluted waters. Moreover, Arantes et al. (2016) reported relatively higher Pb concentrations in a polluted inter-state São Francisco River compared to that observed at Inanda Dam with comparable concentrations observed for Cr and Cd for both populations. Despite metal concentrations in muscle being not conclusive in segregating the two populations, it is evident that the concentration in the muscle does not rapidly respond to the increase in the external water environment. However, analysis of metal concentrations in muscle tissues remains a crucial exercise as muscle is the part consumed by humans.

Despite metal concentrations showing differences between organs, it is evident that metal pollution could be detrimental to aquatic environments. Metals are found in minute concentrations in the water column with much sinking to the bottom sediment. This was also observed in the present study which could be problematic since changes in physico-chemical properties such as pH, water hardness, salinity, temperature etc. could result in metal bioavailability through resuspension back into the water column (Ali et al., 2019; Olayinka-Olagunju et al., 2021). This metal bioavailability through resuspension phenomenon has been a cause for concern in water bodies exhibiting elevated concentrations of metals, as it has implications on the well-being of the environment.

5.3 Metal Concentration Between Organs

The liver and gills are usually the target organs for contaminants due to their role in fish physiology whereas muscle does not have any metabolic role (Ali et al., 2019). In the present study, a descending trend, liver > gill > muscle was observed for Sb, Ni, and Pb at both dams with the inclusive of Cd at Nagle Dam. On the contrary, Cd and Mo at Inanda Dam and Cr and Mo at Nagle Dam exhibited higher concentrations in the muscle compared to other tissues. The liver is a primary site for detoxification whereas gills are in direct contact with the external water environments (Shahjahan et al., 2022), hence, the elevated concentration of metals in both organs is not uncommon. As mentioned earlier, fish muscle usually exhibits low concentrations compared to gills and liver. The liver > gill > muscle trend is comparable to those observed in other related studies in polluted waters (Afandi et al., 2018; Mahboob et al., 2016; Marr et al., 2017). However, other trends such as muscle > liver > gill for Cd, liver > muscle > gill for Cr, and muscle > gill > liver for Mo were observed at Inanda Dam with liver > gill > muscle for Sb, Cd, Ni, and Pb, and muscle > gill > liver for Cr and Mo being observed at Nagle Dam. Shovon et al. (2017) reported an uncommon trend for Cd, Cr, and Mo where the muscle of Labeo rohita and Gibelion catla exhibited higher concentrations compared to the gills. Moreover, Bawuro et al. (2018) reported muscle > gill > liver for Pb and Cd on C. gariepinus and Tilapia zilli.

Despite Inanda and Nagle dam populations exhibiting relatively higher concentrations in the muscle than other tissues for some metals, Sb, Cd, and Cr were within the recommended limits of 1.5 mg/kg (FAO, 1983), 0.3 mg/kg (OJEU, 2006), and 2 mg/kg (MOHSAC, 2006), respectively. On the contrary, Pb concentration exceeded the FAO (2003) and OJEU (2006) limit of 0.02 mg/kg at both dams with the Inanda Dam population exhibiting higher concentration. Moreover, the Ni concentration exceeded the limit of 0.5 mg/kg (UNEP/WHO, 1991) at both dams. Generally, the concentrations of Pb and Ni are concerning and their edibility should be monitored closely.

5.4 Bioaccumulation Factor

The bioaccumulation factor (BAF) may be used to tell whether metals have bioaccumulated or not. In the present study metal concentrations in the water column were below the detection level, hence, the BAF for the water (BAFw) could not be calculated. On the contrary, the bioaccumulation factor for sediment (BAFs) exhibited values ranging from < 1 to > 1. Molybdenum and Ni exhibited BAFs of > 1 in all tissues at Nagle Dam whereas the Inanda Dam population exhibited the BAFs of > 1 for Ni. According to Melake et al. (2023) a BAF of > 1 indicates that metal accumulated more in a tissue whereas a BAF of < 1 suggests that metals accumulated more in sediment and they were not biologically available for fish. The BAFs observed in the present study suggests that Mo and Ni at Nagle Dam may have been biologically available to be taken up by fish. Melake et al. (2023) reported the BAFs of < 1 for Cd, Cr, Ni, and Pb in the muscle of C. gariepinus. In contrast, Olayinka-Olagunju et al. (2021) observed the BAFs > 1 for Pb in the muscle of Parachanna obscura from the Ogbese River. Nevertheless, BAF can be classified into three categories, namely: no probability of accumulation (BAF < 1000), bioaccumulative (1000 < BAF < 5000), and extremely bioaccumulative (BAF > 5000) (Arnot & Gobas, 2006). Based on Arnot and Gobas (2006) classification, there is no probability of accumulation for all metals at both dams. However, this interpretation may be misleading particularly in multi-compartment subjects such as fish where different organs accumulate metals differently depending on their roles, location relative to the medium, and affinity to these metals.

5.5 Metal-Fish Length and Inter-Metal Correlation

Metal accumulation may be influenced by factors such as pH, salinity, temperature, fish size, life history, feeding habits, and fish metabolic rate (Bawuro et al., 2018; Echevarría et al., 2024). According to Jezierska and Witeska (2006), the presence of other metals may also influence bioaccumulation in fish. As a result, metals may display antagonistic or synergistic associations during accumulation. In the present study, inter-metal relationships were observed for Cr–Mo, Cr-Ni, Mo-Ni, and Sb-Pb with fish length showing positive relationships with Cr, Mo, and Ni in the gill and muscle. None of the inter-metal relationships observed in the gill and muscle were observed in the liver with only Sb showing a moderate positive relationship with fish length. Inter-metal relationships observed in the gill and muscle differ from those observed in other related studies (Ahmed et al., 2019; Hashim et al., 2014; Jayaprakash et al., 2015). However, Khan et al. (2023) reported a strong positive relationship for Cr-Ni, Cr-Pb, Ni-Pb in the muscle. Despite liver showing a positive relationship for fish length and Sb, Jezierska and Witeska (2006) and Echevarría et al. (2024) reported that metal concentration may decrease with increasing fish size due to a reduction in feeding as the fish grows, a high metabolic rate in smaller fish and/or dilution factor. The present study has shown not definite trend with regard to inter-metal relationship for all three tissues which affirms that metal affinity, and the physiology, age and sex of fish are among the drivers of metal concentration increase in fish tissues.

5.6 Non-Carcinogenic and Carcinogenic Risk Assessment

Consumption of fish has recently become a cause for concern due to increasing pollution levels in aquatic ecosystems. Despite fish muscle not being the target organ for metals and accumulating lower concentrations compared to other organs such as the liver and gills, these low concentrations are reported to have the probability to induce adverse threats to human health. In the present study, THQs > 1 were observed for Sb, Cr and Pb at both dams with Mo showing a THQ > 1 at Nagle Dam. The Cr and Pb THQs observed in the present study were comparable to those reported further downstream of the river, in the estuarine area of the Indian Ocean for Trachurus trachurus and Chrysoblephus puniceus (Debipersadh et al., 2023), and in the Delta River impacted by industrial, urbanization, and intensive agricultural activities in Ethiopia for Lates niloticus and Oreochromis niloticus (Kotacho et al., 2024). Moreover, Esilaba et al. (2020) observed the Pb THQ of > 1 in Oreochromis niloticus from urban Lake Nakuru in Kenya whereas Lebepe et al. (2020) observed THQ exceeding the acceptable threshold on 1 for Sb on O. mossambicus and L. rosae from water bodies impacted by acid mine drainage. It is evident that anthropogenic activities in the uMgeni River catchment are causing an increase in metal concentration in the system. Other metals such as As, Se, and Hg were also found to exceed acceptable levels for human consumption in numerous fish species at the Inanda and Nagle dams (Misra et al., 2024). On the contrary, fish exhibiting THQs < 1 were was observed in water bodies used for fisheries (Huang et al., 2019; Noman et al., 2022), in protected wetlands (Hossain et al., 2023) and in an urban lake (Han et al., 2021). Nevertheless, low metal concentrations may have a carcinogenic effect if consumed over a long period of time.

In the present study, the CRIs were higher than 10–6 for Cd, Cr and Pb at both dams with the Inanda Dam population exhibiting a relatively higher index compared to Nagle Dam. Similar indices were observed for Trachurus trachurus and Chrysoblephus puniceus in the lower uMgeni River (Debipersadh et al., 2023). However, the CRIs observed at Inanda and Nagle dams were higher than those observed by Kotacho et al. (2024) at the Delta River impacted by industrial and agricultural activities. Moreover, Adegbola et al. (2021) reported CRIs higher than those observed in the present study for Pb and Ni in C. gariepinus at the industrially polluted Ogun and Aleyele rivers. Metals such as Cr, Ni and Pb are showing to be a threat to human health in water bodies impacted by industrial, agricultural and urban activities which suggest that the industrial and urban activities, wastewater works and a landfill could be the explanation for the increased metal concentrations at Inanda Dam. According to (Briffa et al., 2020), metal concentration increases in rivers draining catchments with negligible anthropogenic activities may also be influenced by the geological characteristics of the area. With the Nagle Dam bordered by mountains, the geological characteristics of the area as the driver for the metal increase could not be dismissed.

6 Conclusion

The study explored metal levels at Inanda and Nagle dams to determine the effect of tributaries joining the uMgeni River downstream of the Nagle Dam on the metal increase in the Inanda Dam. The water column exhibited concentrations below detection limits for metals with sediment showing significant concentrations. Moreover, no definite trend was observed for metal concentrations between the two dams, as some metals were higher at Inanda with some being higher at Nagle Dam. Similarly, metal accumulation in fish tissues also showed no definite trend between the two dams and a common trend of liver > gill > muscle was observed for some metals. Nevertheless, it is evident that metal pollution could be a potential threat to the well-being of aquatic life and it should be viewed with concern. Additionally, the muscle of fish exhibited THQs of > 1 for Sb, Cr and Pb with the Nagle Dam population showing a higher THQ for Cr and the Inanda Dam population showing relatively higher THQs for Sb and Pb. Moreover, Sb, Cr and Pb have also shown the probability of carcinogenic effects when consumed over a long period for both populations. Despite polluted tributaries joining the uMgeni River downstream of the Inanda Dam, notable metal concentrations were also observed at Nagle Dam. However, it is evident that the impact of these tributaries manifests in the Inanda Dam. Given the episodic discharge of acidic effluents into the system, it is recommended that the behaviour of metals should be constantly monitored for the sustainability of the well-being of aquatic biota and the safety of fish consumers.