Occurrence, metallo-toxicological interactions, and health risks of selected potentially toxic elements in the River Offin, Ghana

Abstract

Illegal gold mining in the River Offin watershed is causing significant environmental concerns due to the discharge of potentially toxic elements (PTEs) into aquatic ecosystems. This study assessed the occurrence, accumulation, and health implications of arsenic (As), mercury (Hg), and selenium (Se) in water, sediments, and fish from the Offin River, Ghana. Physicochemical parameters, water quality index (WQI), pollution indices, bioconcentration (BCF), biota-sediment accumulation factors (BSAF), and synergistic and antagonistic effects of PTEs were evaluated, and pollution sources were assessed using multivariate statistical analyses. The findings indicated Hg, As, EC, Se, and turbidity concentrations exceeded the guideline limits in water, sediments, and fish samples. The WQI of 7460.609% indicated the water is unsuitable for domestic and agricultural activities. The multivariate analysis revealed that the mercury-based gold mining method influenced the sources, mobility, and transport of contaminants. Based on the health risk assessment, there is a significant likelihood of cancer and other adverse health effects, especially in children, with noncarcinogenic hazard quotients > 1 and carcinogenic risk > 1.00E-4. The synergistic and antagonistic evaluations revealed Se: Hg, Se: As ratios < 1, and Se-HBV < 1, suggesting insufficient selenium protection against Hg and As toxicity in fish, while As: Hg > 1 suggested enhanced Hg toxicity. BCF and BSAF values confirmed that Hg and As accumulate more readily from water to fish than from sediment. This study reinforces the urgent critical for regulatory attention, environmental surveillance, and community awareness to combat illegal mining, minimize PTE exposure and preserve aquatic life and public health in the River Offin basin.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-025-26887-0.

Introduction

Globally, freshwater ecosystems are crucial systems for supporting biodiversity, food security, and human health. Rivers in developing countries increasingly face complex pollution scenarios involving both physicochemical destruction (electrical conductivity (EC), turbidity (turb), total dissolved solids (TDS), pH) and potentially toxic element (PTE) contamination mercury (Hg), arsenic (As), and selenium (Se), threatening aquatic ecosystems and community livelihoods^1–6.

In environmental toxicology, potentially toxic elements (PTEs) such as As, Hg and Se are a category that encompasses both trace elements and heavy metals emphasizing the likelihood of toxicological implications regardless of whether an element is considered essential at lower concentrations^7–10. Exposure to As, Hg, and Se can lead to chronic health issues such as nausea, vomiting, bladder cancer, cardiac disorders, skin irritation, neurotoxicity, and cancer in children and adults^7,11–14. As, Hg and Se are PTEs with significant widespread presence due to their high toxicity through bioaccumulation and environmental persistence and in aquatic ecosystems^1,2,4,6,7,15.

Arsenic is a carcinogenic element harmful to aquatic life, sediments, and water bodies as it biomagnifies through the food chain^7,11,12,14. As has been documented to produce synergistic effects on mercury toxicity, inhibiting its toxicological implications on human health^16–18. Excessive Hg concentrations in freshwater environments pose significant toxicant risks due to their interaction with sulfur residues in proteins, potentially threatening aquatic species and human health¹⁵. While Hg exposure could pose detrimental effects on human health and the environment, its interaction with selenium is complex, as Se may antagonize some of its harmful effects^14,15,19. Selenium (Se) is both an essential element and toxic with a protective role against Hg and As toxicities, which minimizes harmful effects associated with toxicity caused by Hg and As exposure under certain ratios^15,19. The relationship between Hg and Se is complex, but Hg adsorption may be linked to Se’s defensive impact, and its molar ratios (Se:Hg) are crucial for assessing health implications^15,19,20.

The River Offin in Ghana is a vital ecological and socioeconomic resource, providing water, tourism, and fish supplies to the Kumasi metropolis and surrounding areas^11,21–23. Illegal gold mining activities, commonly known as galamsey, release multiple contaminants such as As, Se, and Hg, particularly in large rivers such as Birim, Pra, Offin, and Ankobra in Ghana, causing altered water, food, and environmental quality and increased exposure risks for local communities and aquatic life^1–6,21,24.

Recently prior studies have documented pollution status of individual parts of freshwater ecosystem matrices across Ghana^{5–7,11,21,22,23,25–32} leaving gaps in integrated assessments such as combination of physicochemical, elemental interactions, bioaccumulation degrees, and risk analyses through an interconnected food chain of water fish, soil, and sediment of a particular freshwater system. Even though these local studies have examined physicochemical parameters, including their elemental concentrations, and health risk assessments, the majority of them concentrate only on PTEs or water quality in isolation, paying little attention to elemental concentrations in multiple matrices (water, fish, and sediments), elemental interactions, bioaccumulation, bioconcentration, the antagonistic and synergistic effects of Se, As, and Hg, and the attribution of pollution sources using a variety of methodologies and real-world risk scenarios in biologically relevant compartments.

This study provides the first comprehensive multi-matrix assessment of environmental contamination of the entire ecosystem of the River Offin, Ghana by integrating key physicochemical parameters, potentially toxic elements (PTEs) such as As, Hg, and Se across water, sediment, and fish species. Unlike previous studies that focus on isolated pollutants or single matrices, this research employs a holistically integrated approach to pollution assessment. The novelty of this study lies in the following contributions: first integrated assessment of Se, As, and Hg concentrations in three environmental compartments (water, sediments, and fish species) in the River Offin basin. Utilize the application of selenium health benefit value (HBVSe), and molar ratios (Se: Hg or Se: As), to determine the synergistic and antagonistic interactions of selenium with As and Hg, a rarely studied and underexplored aspect of tropical freshwater ecosystems. Use of bioconcentration (BCF) and bioaccumulation (BSAF) metrics to quantify elemental transport across habitat compartments further complements the ecological risk of fish consumption. Conduct a comprehensive real-time health risk assessment, estimating both the carcinogenic and noncarcinogenic risks for adults and children based on real-time chronic environmental exposure metrics. Application of multivariate statistical tools such as principal component analysis (PCA), and Pearson’s correlation to identify and attribute potential sources of water quality deterioration. The study offers baseline data for environmental managers, policy makers, and sanitation and health authorities to guide remediation strategies, ensure educational awareness on risk communication and sustainable freshwater resource management in River basins. This research contributes to the growing field of environmental geochemistry and aquatic toxicology and presents a replicable framework for risk-based monitoring in other mining-impacted river basins in the Sub-saharan zone.

The objectives of this study are to:

1. Assess the physicochemical status of River Offin, quantify the concentrations of As, Se, and Hg in water, sediment and fish.

2. Evaluate the water quality index, bioconcentration and bioaccumulation factors, and elemental interaction ratios.

3. Assess the health risks for adults and children, pollution status, and potential source origins using multivariate tools.

Materials and methods

Study area

The Offin River happens to be an eastward-flowing waterway in Ghana. It flows through the Tano Offin Reserve in Ghana’s Atwima Mponua District³³. The Offin riverbed is 90 m above mean sea level and has a stretch of 10 km³⁴. The river has witnessed indiscriminate forms of gold mining activities in the river’s sediment. The Offin River serves as an essential source of water and fish to the Kumasi metropolis and its environs through the Barekese and Owabi dams. The Offin River ends as a tributary to the Pra River. Common native fish species found around the Offin River include African filter shrimp, African catfish, Mudfish, Tilapia zillIi, and Blackchin tilapia. The geographical coordinates that were used for this study were 6°39′0″ N and 2°10′0″ W. Figure 1 shows the map of the River Offin.

Fig. 1.

Map of River Offin.

Sampling and pretreatment

Fish, water, and sediment samples were collected in November 2021. Twenty (20) fish samples were collected from local fishermen at varied points along the Offin River stretch (Kyekyewere and Dunkwa) and were used to analyze Hg, Se, and As. The fish species purchased were represented using their sample codes such as F1 (African filter shrimp, Atya gabonensis), F2 (Tilapia zillii, Coptodon zillii), F3 (Mudfish, Clarias agboyiensis), F4 (Blackchin tilapia, Sarotherodon melanotheron) and F5 (African catfish, Clarias gariepinus) before analysis.

The coded fish samples were first frozen with ice in heat insulators and transferred to a freezer in the Ecological laboratory of the University of Ghana. The samples were allowed to stay overnight to completely thaw or defrost. 1000 mg of wet samples of the flesh, intestines, gills, and liver were taken from fish samples. The samples were ground homogenized and put into laboratory bottles with clear identification.

Twenty (20) water samples were randomly collected along the river basin for analysis. Temperature, conductivity, and pH of the water samples were measured and recorded on site. 5 mL of concentrated HNO₃ was added to the water samples before taking it to the lab.

Since gold is mined from the river sediment, twenty (20) sediment samples were collected using an auger using purposive grab sampling. The sediment samples were taken to the laboratory, air-dried, and homogenized. The dried sediment samples were crushed with a clean mortar and pestle and then sieved using a 2 mm sieve. The sieved sediment was placed in ziploc bags and stored for laboratory analysis.

Potentially toxic elements analysis of fish, water, and sediment samples

The 1000 mg samples of each species were later transferred into seven macro-Kjeldahl digestion flasks. 20 mL of concentrated nitric acid (HNO₃) was added to all samples and topped with about 20 mL of distilled water (D₂O). The samples were boiled to reduce the volume to 20 mL. The digestates were cooled, and 10 mL of concentrated sulphuric acid was added to each and boiled again. 10 mL of nitric acid was added to prevent the contents from blackening. After blackening had ceased, the contents were continuously heated until all white fumes evolved. At this stage, the digestates were cooled, and 10 mL of saturated ammonium oxalate solution was added to each and again boiled until copious white fumes were seen. Oxalate treatment effectively removed yellow coloration due to nitro compounds and fats, ensuring a colorless solution for metal assay, and eliminating nitric acid trace, crucial for accurate metal analysis. The digestates were then allowed to cool and filtered using Whatman No. 42. A volume of 15 mL of each digestate was filtered and placed in waiting bottles for final analysis, ensuring their quality and safety. Blank samples were also prepared for similar treatment without fish samples.

A mass of 1 g of sediment sample and 20 mL of a 1:1:1 mixture of HNO₃: HCl: D₂O mixture were heated to boiling at 90 ℃ for 1 h on a Kjeldahl digestion apparatus. The digestate was filtered with Whatman no.42 and topped up to the 50 mL mark using 35% HCl before the analysis.

A mass of 50 mL of the acidified water sample was measured and 10 mL of HNO₃ was added and heated to boiling at 150 ℃ for 1 h to reduce to 20 mL. The digestate was cooled and filtered using a Whatman no. 42 filter to remove suspended particles and topped up to the 50 mL mark using D₂O.

The cold vapor generation method for total Hg determination was performed using 0.3% (w/v) of sodium borohydride (NaBH₄) as a reductive solution to determine total mercury (THg). 5 M hydrochloric acid (HCl) was used as a carrier solution. Hydride generation was performed with a 0.6% (w/v) NaBH₄ reductive solution for both total Se and As determination. This solution was prepared daily or more frequently when required. 6 M HCl was used as the carrier solution.

Total mercury in the fish samples was measured using the Cold vapor atomic absorption spectrometry (CV-AAS) coupled with an Automatic mercury analyzer (HG5000, Japan) method, while total arsenic and selenium were measured using the hydride generation atomic absorption spectrometer (HG-AAS, AA 240 FS, USA).

Quality assurance

Reagent blanks, replication samples, pre-digestion spikes, and post-digestion spikes were among the quality assurance samples examined. Additionally, the technique’s limits of detection (As, Se = 0.001, and Hg = 0.0001 mg/L), slope (0.99), and rate of recovery (97%) were performed. Various quantities of standard mercury solutions were added to fish samples to verify the analytical methodology and conduct recovery repeatability experiments. To determine the levels of Hg, Se, and As in the collected samples, the same digestion process was used for triplicate tests, spiked samples, and controls in each run.

Water quality index

In this investigation, the weight arithmetic water quality index (WQI) was used to ascertain if surface water samples from the River Offin were appropriate for a certain purpose based on the assigned weight (wi) and relative weight (Wi) estimators in Eq. (1)^3,7,35. The WQI was computed using Eqs. (1) and (2). This approach evaluated the WQI based on a rating scale (1–5) indicating less to high impactful parameters to health. The standard (S_i) values for the various parameters were used^3,35–37. The relative weight (Wi) was determined from Eq. (3) using the modified suggested limits³. Depending on the degree of risk to human health, each parameter was given a weight between 1 and 5 (Table 1). All variables required for the estimation of the WQI of River offin has been presented in Table 1. Excellent water (< 50), good water (50–100), poor water (100–200), extremely poor water (200–300), and river water that is unfit for human consumption (> 300) are the five (5) categorized measures that comprise the water quality index^3,7.

1

2

3

where WQI = Water quality index, q_{i =} quality rating, W_i = relative weight, w_i = assigned weight of each parameter, C_i = mean concentration of each parameter, S_i = drinking water standards for each parameter, n = number of parameters.

Table 1.

Water quality index of the River Offin.

Parameter	Standard (Si)	Assigned Weightage (wi)	Relative Weight (Wi)
pH	8.5	4	0.220
Turbidity (NTU)	5	5	0.220
EC (μs/cm)	400	2	0.170
Hg (mg/L)	0.001	5	0.130
As (mg/L)	0.01	5	0.130
Se (mg/L)	0.01	4	0.130
		Σ wi = 25	Σ Wi = 1

Human health risk assessment

The human health risk assessment examined the potential for potentially toxic elements in fish, water, and sediments at the Offin River sampling site to initiate adverse health consequences in adults and children by ingestion¹¹. Ingestion was the most critical exposure pathway for PTEs in all samples affecting human health^3,6,11,12. The USEPA categorizes As, Se, and Hg as non-carcinogenic PTEs, while As is classified as carcinogenic^11,35. To ascertain the human health risks, the estimated dietary intake, non-carcinogenic risks, and carcinogenic risks were computed.

Average daily intake

To evaluate the exposures to public health posed by PTEs in sediment (Eq. 4), water (Eq. 5), and fish (Eq. 6) through ingestion the average daily intake (ADI) (mg/kg/day) of PTE was calculated. The exposure parameters used for the calculation of the health risk associated with the consumption of sediments^11,38, water^3,12 and fish^6,39 in adults and children are shown in Table 2.

4

5

6

Table 2.

Exposure parameters for estimating the health associated with the ingestion of fish, water, and sediments in adults and children.

Water variables	Description	Units	Children	Adults	Reference
Cwater	Concentration of water	mg/L	This study	This study	This study
WIR	water intake rate	L/day	1.8	3	3,42
EF	Exposure frequency	days/year	365	365	3
ED	Exposure Duration	year	10	40	3
AT	Average Time	days	3650	14,600	3
BW	Body Weight	kg	15	70	3
Sediment variables
Csediment	Concentration of sediment	mg/kg	This study	This study	This study
SIR	sediment intake rate	mg/day	200	100	11
EF	Exposure frequency	days/year	350	250	11,45
ED	Exposure Duration	year	30	6	11
AT	Average Time	days	10,950	2190	11
BW	Body Weight	kg	15	70	11
CF	Unity conversion factor	kg/mg	10–6	10–6	11
Fish variables
Cfish	Concentration of fish	mg/kg	This study	This study	This study
FIR	Fish intake rate	mg/day	200	100	6
EF	Exposure frequency	days/year	365	365	6
ED	Exposure Duration	year	30	6	6
AT	Average Time	days	10,950	2190	6
BW	Body Weight	kg	15	70	6
CF	conversion factor	kg/mg	0.208	0.208	39

Non-carcinogenic risk

The unitless hazard quotient (HQ) was utilized to estimate the non-carcinogenic risk of consuming PTE-contaminated fish, sediment, and water at which no adverse health effects are expected. The HQ was calculated as the fractional ratio of PTE ADI through ingestion to the reference dose (RfD) for each PTE using Eq. (7)¹¹. The chronic oral reference doses (RfD) (mg/kg/day) for Hg (0.0001), As (0.0003), and Se (0.005)^40–43. The cumulative hazard index (HI), was used to assess the combined non-carcinogenic adverse health risk for all PTEs (HQ) in all samples. The HI was estimated using Eq. (8). An HI greater than 1 implies an elevated likelihood of a combined detrimental non-carcinogenic effects of all PTE on children, and adults, who consume fish, water, and sediment^3,6,11,12. All variables for the estimation of the HQ and HI are presented in Table 2.

7

8

Carcinogenic risk

Carcinogenic risk (CR) is the greater likelihood that adults or children are susceptible to signs and symptoms of cancer as a result of being exposed to cancer-causing PTE during their lifetime¹¹. The CR for the As was calculated using Eq. (9).

9

The cancer slope factor (CSF) estimates the greater risk of cancer in adults and children associated with lifetime exposure to a carcinogenic substance (As)¹¹. The SF for As is 1.5^6,40,43. A CR unit of less than 1.00E-6 for a single PTE indicates less likelihood of cancerous developments in any part of the exposed body. However, a risk greater than 1.00E-4 is considered highly unacceptable and a cause for alarm. All variables for the estimation of the CR are presented in Table 2.

Aquatic bioaccumulation model

The aquatic bioaccumulation model (ABM) was used to determine the bioaccumulation potential of fish organisms through the water or the sediment as a carrier or medium of exposure to increase the concentrations of PTEs in fish. The ABM can also infer which medium greatly impacts the fish accumulation of PTEs in the river environment. To determine the ABM, bioaccumulation and bioconcentration factors were used^12,44.

Bioconcentration factor

The bioconcentration factor (BCF) is a metric used in aquatic toxicology to quantify the concentration of a PTE or other pollutant in an organism (fish or aquatic species) to its ecosystem (river)^12,44. The BCF shows how the occurrence of PTE in the surrounding water of the fish affects the concentrations of PTE in the fish. The proportion of a specific PTE in water to the ratio of PTE in an organism is known as the BCF of PTE. Equation 10 will be used to determine the BCF for fish in the Offin River^12,44. BCF values (L/kg) > 1 indicate that the concentration in the fish is greater than that of the water from which the PTE was taken or it indicates significant accumulation of PTE in the organism (fish) while BCF < 1 indicates that the PTE is accumulating at a lower concentration in the fish than in the water implying minimal accumulation^12,44.

10

Where, _fish refers to the mean concentration of the fish samples from River Offin, refers to the mean concentration of the water samples from River Offin.

Biota-sediment accumulation factor

Biota-sediment accumulation factor (BSAF) is used to ascertain the rate of absorption and excretion rate of a chemical substance (PTE) by an organism (fish) by quantifying the ratio of concentrations of PTEs in sediments to that of the organism^12,44. BSAF determines how the occurrence of PTE in sediments affects the build-up of concentrations of PTE in fish. The BSAF is calculated using Eq. (11).

11

_fish refers to the mean concentration of the fish samples from River Offin.

refers to the mean concentration of the sediment samples from River Offin.

BSAF values > 1 indicate that the concentration in the fish is greater than that of the sediment from which the PTE was taken or it indicates a significant accumulation of PTE in the organism (fish) while BSAF < 1 indicates that the PTE is accumulating at a lower concentration in the fish than in the sediment implying minimal accumulation^12,44.

Estimation of selenium to mercury and selenium to arsenic molar concentrations in fish

The molar ratio of selenium to mercury and selenium to arsenic was assessed to identify the synergistic effect (enhancing toxicity of mercury) and the antagonistic effect (reducing the toxicity of mercury and arsenic) in the fish species. First, the molar PTE unit was estimated using the concentration of an individual PTE such as Se, As, and Hg (mg/kg ww) in the fish were divided by their molar masses 78.963, 74.922 and 200.590 g/mol respectively as utilized in Eq. (12). The Se: Hg, Se: As, and As: Hg molar ratios were calculated using Eq. (13) to define the synergistic and antagonistic effects of Se in this study. All molar ratio average values reported in this study were averaged from fish specimen values for each species.

12

13

Where, C PTE refers to the concentration of ith toxic element, mg/kg (Hg, As and Se), Mm PTE refers to the molar mass (g/mol) of ith toxic element (Hg, As, and Se).

An Se: Hg; or Se: As > 1 refers to excess Se in the system to prevent Hg and As toxicity and consequently demonstrates an antagonistic reaction on Hg and As whereby a Se: Hg; Se: As < 1 signals Hg or As toxicity due to Se deficiency^17,45.

An As: Hg > 1 however shows excess As in the system which will lead to a synergistic influence on Hg toxicity in the fish. A Se: Hg molar ratio ≥ 1 shows that there are adequate levels of Se to antagonize the toxicity of Hg^17,45.

Selenium health benefit value (HBV_Se)

The health risk benefit value selenium (HBV_Se) estimations take into account the body’s biochemical use of Se, which is required for tasks such as the production of selenoproteins involved in various beneficial biological functions^17,18,45.

A more recent measure of protection based on Se on Hg and Se on As levels in fish is done using the Se health benefit value (HBV_Se) in Eqs. (14) and (15)^17,46,47. Equation 16 was used to estimate the cumulative action of Se production on risks associated with the consumption of As and Hg contaminated fish.

14

15

16

Where, Se, As, and Hg are molar amounts, A THBV_Se value > 0 suggests the presence of Se in excess for the body’s biochemical demand and is, therefore, considered protective; a THBV_Se value < 0 is considered risky, indicating that inhibition or sequestration of As and Hg by Se is unlikely^17,46.

Pollution potential indices

The assessment of pollution potential indices in the environmental matrices (sediment, water and fish) was subjected to a series of measurements to ascertain the extent of combined pollution effects of arsenic, mercury, and selenium in fishes, water, and sediment. The pollution potential assessments used were the contamination degree, heavy metal evaluation index, and heavy metal pollution index for all environmental matrices analyzed in this study.

Heavy metal evaluation index (HMEI)

The heavy metal evaluation index (HMEI) was calculated using Eq. (17)^35,48.

17

Where, Mi refers to the measured concentration of PTE in riverine (µg/L), sediments (µg/kg), and fish samples (µg/kg), MACi refers to the maximum permissible concentration (µg/L) of each PTE. The MACi; in water samples µg/L for Hg (1), As,, and Se (10)^36,37, in fish samples, µg/kg, Hg (2), As (25) and Se (50)^39,49, and sediment samples, ug/kg Hg (300), As (12,000) and Se (1000)^2,50–52.

The HMEI has three classifications for categorizing river water, fish, and sediment : low (< 10), moderate (10–20), and high pollution and warning threshold (> 20)^35,48.

Degree of contamination

The degree of contamination (C_d) is the cumulative hazardous impact of PTE on river water, sediments, and fish which was ascertained with Eq. (18)³⁵.

18

where, Cfi is the contamination factor for the ith PTE, Mi is the measured concentration of toxic elements in riverine samples (µg/L), sediment, and fish samples (µg/kg), MACi is the maximum permissible concentration (µg/L) of each potentially toxic element, The MACi; in water samples µg/L for Hg (1), As and Se (10)^36,37, in fish samples, µg/kg, Hg (2), As (25) and Se (50)^39,49, and sediment samples, ug/kg Hg (300), As (12,000) and Se (1000)^2,50–52, Cdeg is classified as follows: For low < 1, moderate within 1–3, and > 3 extremely high contamination levels of PTE in river water, sediment, and fish³⁵.

Data analysis

Microsoft Excel was employed to determine the descriptive statistics such as the standard deviations, one sample t-test, maximum, minimum, mean values, and upper and lower confidence limits of the data. The upper and lower confidence limits provide a range of values that are likely to contain the standard limit value at an error rate of 5%. The one-sample t-test was used to ascertain the statistically significant difference in the summarized datasets and to compare the concentrations of quality parameters to their acceptable permissible limits. Pearson’s correlation was used to establish a relationship between the physicochemical parameters and the PTE in the sediment, fish, and water samples. Minitab 20.3 software was used to analyze the principal component analysis (PCA) to determine possible sources of contamination.

Results and Discussion

Concentration of toxic elements and physicochemical parameters in Offin River

The concentration of arsenic (As), mercury (Hg), selenium (Se), pH, electrical conductivity, and turbidity of water samples from River Offin are presented in Table 3.

Table 3.

The concentration of PTEs, water quality index and physicochemical parameters in water, sediments and fish from the Offin River.

Parameter	Min	Max	Mean	StDev	WHO	t-crit	t-stats	p-value	UCL	LCL
WATER
Hg	0.003	0.004	0.004	0	0.001	1.833	55.233	5.26E − 13	0.004	0.003
As	0.013	0.019	0.016	0.002	0.01	1.833	9.307	3.24E − 06	0.018	0.015
Se	0.012	0.017	0.015	0.002	0.01	1.833	8.822	1.00E − 05	0.016	0.014
pH	7.1	7.3	7.16	0.070	6.5–8.5	1.833	− 60.604	4.57E − 13	7.21	7.11
EC	480	530	497	14.546	400	1.833	21	5.90E − 09	507.006	486.194
Turbidity	1576	2015	1825	133.497	5	1.833	43.119	4.84E − 12	1920.798	1729.802
WQI	7460.609
SEDIMENT
Hg	2	2.3	2.18	0.103	0.3	1.833	57.563	3.63E − 13	2.254	2.106
As	1	1.3	1.14	0.107	12	1.833	− 319.473	7.34E − 20	1.217	1.063
Se	1.7	2	1.86	0.117	1	1.833	23.169	1.24E − 09	1.944	1.776
pH	7.1	7.3	7.19	0.088	7.5	1.833	− 11.196	6.94E − 07	7.253	7.127
EC(µS/cm)	360	380	371	7.008	–	1.833	167.411	2.46E − 17	376.013	365.987
FISH
Hg	0.412	0.642	0.517	0.109	0.002	2.132	10.524	0.0002	0.652	0.381
Se	0.02	0.043	0.028	0.01	0.05	2.132	− 5.212	0.0030	0.04	0.016
As	2.31	5.641	4.196	1.334	0.025	2.132	6.991	0.0010	5.852	2.539

The dissolved Hg concentration varied between 0.003 and 0.004 mg/L with a mean of 0.004 mg/L. Based on the statistical analysis, the mean Hg of 0.004 mg/L was statistically higher than (+ ve t-stats (+ 55.233), t-critical (1.833) < t-stats (+ 55.233)) the recommended threshold of 0.001 mg/L^36,53. In addition, the one sample t-test indicated Hg concentrations were statistically different (p = 5.26E-13 < 0.05) among all water samples investigated, between the sample mean and their Hg standard value. The 95% UCL (0.004 mg/L) and LCL (0.003 mg/L) also indicate statistically greater than their standard value of 0.001 mg/L. The high Hg concentration in water may pose potential neurotoxicity and kidney problems in children and adults^6,22,54. In contrast to our study, the mean Hg concentration (0.004 mg/L) was than higher Hg concentrations of 0.061 mg/L and 0.009 mg/L in Ankobra and Pra Basins in Ghana⁶, indicating elevated concentrations of Hg due to differences in amalgamation practices from alluvial mining, organic fungicides and organic herbicides application near the river banks respectively.

Arsenic is a globally known carcinogenic element. The dissolved As concentration ranged between 0.013 to 0.014 mg/L with a mean of 0.016 mg/L. Based on the statistical analysis, the mean As of 0.016 ± 0.002 mg/L was statistically higher than (+ ve t-stats (+ 9.307), t-critical (1.833) < t-stats (+ 9.307)) the recommended threshold of 0.01 mg/L^36,53. In addition, the one sample t-test showed that As concentrations were statistically different (p = 3.24E-06 < 0.05) among all water samples investigated, between the sample mean and their As standard value.

The 95% UCL (0.018 mg/L) and LCL (0.015 mg/L) of As concentration also indicate statistically greater than their standard value of 0.010 mg/L. The high As concentration in water may contribute to potential skin cancer, pigmentation, kidney, vascular, and brain disorders in children and adults^6,22,54. In contrast to our study, the mean As concentration (0.016 mg/L) was similar to As concentrations of 0.021 mg/L in Mantaro River in Central Andes of Peru⁵⁵ and 0.010 mg/L in Coaplex stream, South Africa⁵⁶ and less than the As concentration of 0.04 mg/L in River Pra and Ankobra basin, Ghana. The variation in As concentrations may be attributed to the extent of runoffs from; gold ore mining, tin smelting, and metallurgical waste used during illegal mining processes as well as industrial inferences⁵⁵.

Selenium (Se) is an essential trace element with significant protective implications or human health. The Se concentration ranged between 0.012 to 0.017 mg/L with a mean of 0.015 ± 0.002 mg/L. Based on the statistical analysis, the mean Se of 0.015 mg/L was statistically greater than (+ ve t-stats (+ 8.822), t-critical (1.833) < t-stats (+ 8.822)) the recommended acceptable level of 0.01 mg/L ^36,53. In addition, the one sample t-test showed that Se concentrations were statistically different (p = 4.57E-13 < 0.05) among all water samples investigated, between the sample mean and their Se standard value. The 95% UCL (0.016 mg/L) and LCL (0.014 mg/L) of Se concentration were statistically greater than their standard value of 0.010 mg/L. The Se concentration in this may assist in the reduction of certain cancers, cardiovascular disorders, and improve reproductive health while its deficiency may be associated with viral infections, cognitive decline and autoimmune diseases ^57–59. In contrast to our study, the mean Se concentration (0.015 mg/L) was less than the Se concentrations of 2.58E-03 mg/L in Bitter Lakes Egypt ¹⁵, 2.79E-03 mg/L in Mokolo River, South Africa ⁶⁰ and 7.70E-05 mg/L in Nakambe watershed reservoirs, Burkina Faso ⁶¹. The variation in Se concentrations may be due to efficient natural attenuation mechanisms and laws that restrict selenium discharges ⁶⁰, attributed to poor bioavailability, low light intensity, salinity, and lack of weathering of Se rocks, soil, selenium sources and the dilution effects of Pra and Ankobra Rivers ^15,62.

The pH is measure of the acidity, and alkaline status of the River water. The pH values ranged from 7.100 to 7.300 with a mean of 7.160 ± 0.070. Based on the statistical analysis, the mean pH of 7.160 was statistically less than (-ve t-stats (-60.604), t-critical (1.833) > t-stats (-60.604)) the recommended acceptable range of 6.5—8.5 ^36,53. In addition, the one sample t-test showed that pH concentrations were statistically different (p = 4.57E-13 < 0.05) among all water samples investigated, between the sample mean and their pH standard value. The 95% UCL (7.210) and LCL (7.110) pH levels were within their standard value of 6.5–8.5. The pH of the River water indicated neutral levels within recommended limits. A neutral pH may not be indicative of clean water as pH may not be the only variable for measuring the quality status of the river. In contrast to our study, the mean pH (7.10) was similar to the neutral pH level of 7.13 in Coaplex stream, South Africa ⁵⁶, and less than the alkaline pH level of 8.09 in Meghna River, Bangladesh ⁶³. The variation in pH may be due to industrial discharge of acidic effluents, dissolved carbonates instrumental deviation, sampling location, and time of analysis which may be indicative that anthropogenic processes presently does not influence alkalinity or acidity content of the river ³.

The electrical conductivity (EC) is an indirect measure of the salinity of the River water. The EC values ranged from 480 to 530 µs/cm with a mean of 497 ± 14.546 µs/cm. Based on the statistical analysis, the mean EC of 497 µs/cm was statistically greater than (+ ve t-stats (21.00), t-critical (1.833) < t-stats (+ 21.00)) the recommended limit of 400 µs/cm for drinking ^3,36,53. In addition, the one sample t-test showed that EC concentrations were statistically different (p = 5.90E-09 < 0.05) among all water samples investigated, between the sample mean and their EC permissible limit. The 95% UCL (507.006 µs/cm) and LCL (486.194 µs/cm ) EC levels were statistically greater their standard value of 400 µs/cm ^3,36,53. The mean EC values of Offin River water less than 700 µs/cm indicated low levels of dissolved mineral content, lower ionic concentration and poor mineralization ^3,35 which may posess suitable quality for irrigation purposes ⁶⁴. In comparison to this study, the mean EC (497 µs/cm) was similar to the EC concentration of 650 µs/cm in Coaplex stream, South Africa ⁵⁶, and 530 µs/cm in Sisa River, Ghana ³⁵. The variation in EC may be due to differences in ionic charge species, inorganic salt, organic matter, instrumental deviation, sampling location and time of analysis may have influenced the results ³.

The turbidity is a physical evaluation of the clearness, transparency and visual appeal status of the River water for drinking or treatment. The turbidity values ranged from 1576 to 2015 NTU with a mean of 1825 ± 133.497 NTU. Based on the statistical analysis, the mean turbidity of 1825 NTU was statistically greater than (+ ve t-stats (43.119), t-critical (1.833) < t-stats (+ 43.119)) the recommended limit of 5.000 NTU for drinking and domestic purposes ^3,36,53. In addition, the one sample t-test showed that turbidity levels were statistically different (p = 4.84E-12 < 0.05) among all water samples investigated, between the sample mean and their turbidity permissible limit. The 95% UCL (1920.798 NTU) and LCL (1729.798 NTU) turbidity levels were statistically greater than their standard value of 5 NTU ^3,36,53,65. The mean turbidity values obtained in this study showed that Offin River water was very dirty and brownish due to the high turbid values indicating that the river water was unacceptable for drinking. In comparison to this study, the mean turbidity (1825 NTU) from Offin river water was greater than to the turbidity concentration of 57.40 NTU in in Sisa River, Ghana ³⁵, 80.73 NTU in Karnaphuli River, Bangladesh ⁶⁶, and 0.88 NTU in Glacial lakes, Romania ⁶⁷. Differences in excessive soil dredging activities for gold ore, clumps of dirt and soil particles, presence of sludge, organic matter, hydroxide or aluminium and iron from mining tools waste and dumps, colloidal bacteria might have accounted for the variation in the turbidity of surface water in the study ^3,67.

Concentration of toxic elements and physicochemical parameters in sediments

The concentration of arsenic (As), mercury (Hg), selenium (Se), pH, and electrical conductivity, of sediment samples from the Offin riverbed are presented in Table 3.

The pH level in Offin river sediments is a crucial assessment to determine the retention or sorption, mobility or transport and the bioavailability of nutrients, PTEs and other pollutants hanging on sediments ^2,68. The pH values ranged from 7.100 to 7.400 with a mean of 7.160 ± 0.088 in sediment samples. Based on the statistical analysis, the mean pH of 7.190 was within (-ve t-stats (-11.196), t-critical (1.833) > t-stats (-11.196)) the recommended acceptable range of 6.5—8.5 ^36,53. In addition, the one sample t-test showed that pH concentrations were statistically different (p = 6.94E-07 < 0.05) among all water samples investigated, between the sample mean and their pH standard value. The 95% UCL (7.253) and LCL (7.127) pH levels were within their standard value of 7.5 ⁶⁸. The pH of River water indicated neutral levels within recommended limits. A neutral pH may not be indicative of clean water as pH may not be the only variable for measuring the quality status of the river. Also the neutral pH (7.160) in this study does not favor PTEs, nutrients and other pollutants retention, indicates that PTEs, nutrients and other pollutants are bioavailable to phytoplankton and promote their desorption into the water system. In contrast to our study, the mean pH (7.160) was similar to the neutral pH level of 7.13 in Coaplex stream, South Africa ⁵⁶, and 7.2 in mining sediments, Ghana ⁶⁸, while being less than the alkaline pH level; of 8.09 in Meghna River, Bangladesh ⁶³, 9.6 in Wuliangsuhai lake sediments, Mongolia ⁶⁹. The variation in pH may be due to industrial discharge of acidic effluents, dissolved carbonates instrumental deviation, sampling location and time of analysis which may be indicative that anthropogenic processes presently does not influence alkalinity or acidity content of the river ^2,3,68.

The electrical conductivity (EC) is indirect measure of the salinity of the River water. The EC values ranged from 360 to 380 µs/cm with a mean of 371 ± 7.008 µs/cm. In addition, the one sample t-test showed that EC concentrations were statistically different (p = 2.46E-17 < 0.05) among all water samples investigated, and between the sample mean. The mean EC values of Offin river sediments (371 µs/cm) was less than the EC of 1341.5 µs/cm in Wuliangshai lake sediment in Mongolia ⁶⁹ and EC of 725 µs/cm in Shallow lakes open water sediments, Austria ⁷⁰. The variation in EC may be due to differences in salinity concentration, clay minerals from soil dredging activities, mining effluents after gold ore rock harvesting, agricultural runoffs, dissolution of sulphates and carbonates may have influenced the results ^3,69.

The Hg concentration ranged between 2.000 to 2.300 mg/kg with a mean of 2.180 ± 0.103 mg/kg in sediment samples. Based on the statistical analysis, the mean Hg of 2.300 mg/kg was statistically higher than (+ ve t-stats (+ 57.563), t-critical (1.833) < t-stats (+ 57.563)) the recommended threshold of 0.300 mg/L in sediments ^2,51. In addition, the one sample t-test showed Hg concentrations were statistically different (p = 3.63E-13 < 0.05) among all sediment samples investigated, between the sample mean and their Hg standard value.

The 95% UCL (2.254 mg/kg) and LCL (1.063 mg/kg) also indicate statistically greater than their standard value of 0.300 mg/L in sediment samples ². The high Hg concentration in sediments may pose high accumulation and bioavailability in phytoplankton and aquatic life ^2,68,71. In comparison to our study, the mean Hg concentration of 2.180 mg/kg in Offin River sediments was than higher average Hg concentrations in Bonsa River, Ghana (1.080 mg/kg) ⁷¹, Nyam River (0.15 mg/kg) and Atiwa River (0.07 mg/kg), Ghana ⁶⁸, Patuakhali coastal sediments, Bangladesh (0.06 mg/kg) ⁷², Serbian Rivers and lakes sediemnts, Serbia (0.183 mg/kg) ⁷³ and South Morava River, Southeastern Serbia (0.600 mg/kg) ⁷⁴, indicating poor quality sediments. The differences in bioavailability of Hg species in the sediment samples indicates elevated concentrations of Hg may be attributed to frequency of gold ore amalgamation that is adsorbed onto the river bed, and other non-point sources.

The As concentration ranged between 1.000 to 1.300 mg/kg with a mean of 1.140 ± 0.107 mg/kg in sediment samples. Based on the statistical analysis, the mean As of 1.140 mg/kg was statistically less than (-ve t-stats (-319.473), t-critical (1.833) > t-stats (-319.473)) the recommended acceptable level of 12 mg/kg ^2,51. In addition, the one sample t-test showed that As concentrations were statistically different (p = 7.34E-20 < 0.05) among all sediment samples investigated, the sample mean and their As standard value. The 95% UCL (1.217 mg/kg) and LCL (1.063 mg/kg) As concentrations also indicate that they are statistically less than their standard value of 12.000 mg/kg ^2,51. The high As concentration in water may contribute to biomagnification and bioaccumulation in phytoplankton, fish, and the food chain. In comparison to our study, the mean As concentration of 1.140 mg/kg in Offin River sediments was higher than the average As concentrations in Bonsa River, Ghana (0.74 mg/kg) ⁷¹, Patuakhali coastal sediments, Bangladesh (0.021 mg/kg) ⁷², and coastal belt of Kerala, India ⁷⁵. In addition, the mean As concentration was less than the average As concentration in Nyam River (397.080 mg/kg) and Atiwa River (42.860 mg/kg), Ghana ⁶⁸, and Serbian Rivers and lakes sediments, Serbia (19.000 mg/kg) ⁷³. The differences in As concentration in sediments may be attributed to agricultural herbicides and pesticide runoffs, seasonal trends, erosion, dumped chemical munition containing arsenic-based Chemical Warfare Agents, rusting of mine pans, and As-rich mine tailings, which may have impacted the results ^2,68,76,77.

The Se concentration ranged from 1.700 to 2.000 mg/kg with a mean of 1.860 ± 0.117 mg/kg in sediment samples. Based on the statistical analysis, the mean Se of 1.860 mg/kg was statistically greater than (+ ve t-stats (+ 23.169), t-critical (1.833) > t-stats (+ 23.169)) the recommended acceptable level of 1.000 mg/kg ⁵¹. In addition, the one sample t-test showed that Se concentrations were statistically different (p = 1.24E-09 < 0.05) among all sediment samples investigated, between the sample mean and their Se standard value. The 95% UCL (1.944 mg/kg) and LCL (1.776 mg/kg) Se concentrations also indicate that they are statistically greater than their standard value of 1.000 mg/kg ⁵¹. The Se concentration in this study may interfere with the reproduction cycle of egg-laying vertebrates, larvae and developmental abnormalities and mortality, bioaccumulate, and induce teratogenic implications ⁷⁸. The mean Se concentration (1.860 mg/kg) in River Offin sediments was higher than the average Se concentrations in Serbian rivers and lakes sediments, Serbia (0.34 mg/kg) ⁷³, Tapajos River sediments, Brazil (0.664 mg/kg) ⁷⁹, Kagoshima Bay, Japan (0.54 mg/kg) ⁸⁰ and less than the mine effluents in sediments, Canada (321 mg/kg) ⁸¹. The variation in results may have been impacted by differences in marine biological materials, geothermal activity supplies, sedimentary rocks and shales, gold ore mining effluents enhanced by acid mine drainage, fossil fuel refinement, and irrigation of Se-rich rocks ^78–80.

Concentration of toxic elements parameters in fishes

The concentration of arsenic (As), mercury (Hg), selenium (Se), in fish samples from River Offin are presented in Table 3.

The Hg concentration ranged between 0.412 to 0.642 mg/kg with a mean of 0.517 ± 0.109 mg/kg in fish samples. Based on the statistical analysis, the mean Hg of 0.517 mg/kg was statistically higher than (+ ve t-stats (+ 16.524), t-critical (2.132) < t-stats (+ 16.524)) the recommended threshold of 0.002 mg/L in fishes ^6,39. In addition, the one sample t-test showed Hg concentrations were statistically different (p = 0.0002 < 0.05) among all fish samples investigated, between the sample mean and their Hg permissible limit. The 95% UCL (0.652 mg/kg) and LCL (0.381 mg/kg) of Hg in fish species was also statistically greater than their standard value of 0.002 mg/L ^6,39. The high Hg concentration in fishes may pose induce toxicity in fishes and consumers who rely on the fish for food.

In contrast to our study, other authors reported higher Hg concentrations; of 0.60 mg/kg in fish species from Ankobra and Pra Basins in Ghana ⁶, 0.137 mg/kg in fish from Erbul city markets, Iraq ⁸², and 0.063 mg/kg in freshwater fish from Poland ⁸³, indicating higher likelihood of Hg associated health consequences in the consumption of fish. The differences in bioavailability of Hg species in the fish samples may be attributed primarily to illegal mercury-based mining activities within the catchment ⁸⁴.

The As concentration ranged between 0.02 to 0.0043 mg/kg with a mean of 0.028 ± 0.010 mg/kg in fish samples. Based on the statistical analysis, the mean As of 0.028 mg/kg was statistically less than (-ve t-stats (-5.212), t-critical (2.132) > t-stats (-5.212)) the recommended acceptable level of 0.05 mg/kg ^6,39. In addition, the one sample t-test showed that As concentrations were statistically different (p = 0.003 < 0.05) among all fish samples investigated, the sample mean and their As standard value. The 95% UCL (0.040 mg/kg) and LCL (0.016 mg/kg) As concentrations in fish species also indicate that they are statistically less than their standard value of 0.05 mg/kg ^6,39. In comparison to our study, the mean As concentration of 0.028 mg/kg in fish was less than the As concentrations in Oman sea, Iran (1.44 mg/kg) ¹⁷, and Nakambe watershed, Burkina Faso (0.080 mg/kg) ⁶¹. The differences in As concentration in fish may be attributed to fish species, As-uptake rate of the fish, anthropogenic sources and parts of the fish analyzed may have impacted the results.

The Se concentration ranged from 2.310 to 5.641 mg/kg with a mean of 4.196 ± 1.334 mg/kg in fish samples. Based on the statistical analysis, the mean Se of 4.196 mg/kg was statistically greater than (+ ve t-stats (+ 6.991), t-critical (2.321) > t-stats (+ 6.991)) the recommended acceptable level of 0.025 mg/kg ^42,85,86. In addition, the one sample t-test showed that Se concentrations were statistically different (p = 0.001 < 0.05) among all fish samples investigated, between the sample mean and their Se standard value. The 95% UCL (5.852 mg/kg) and LCL (2.539 mg/kg) Se concentrations also indicated that they are statistically greater than their standard value of 0.025 mg/kg ^42,85,86.

The high Se concentration in this study may cause teratogenic defects, such as gradual deformities of the head, mouth, and fins in fish after prolonged exposure ⁵⁷. The mean Se concentration (4.196 mg/kg) in fish was higher than the Se concentrations in the Oman Sea, Iran (1.38 mg/kg), Nakambe watershed, Burkina Faso (0.33 mg/kg) ⁶¹, Turkish coast, Turkey (1.73 mg/kg) ¹⁴, Tapajos River, Brazil (0.41 mg/kg) ²⁴, Coastal lagoons, Gulf of California (2.88 mg/kg) ⁴⁷, and higher than Se concentrations in Balearic Mediterranean sea, Spain (51.776 mg/kg) ⁸⁷. The variability in results may have been impacted by differences in fish species, anthropogenic activities and the water source.

Water quality index

The water quality index (WQI) of the Offin River was 7460.609% indicating unsuitable water for domestic, irrigation, and industrial activities (Table 3). The unsuitable water quality status in the Offin River was similar to the unsuitable water status in the Ghaghara River River in India ⁸⁸. The turbidity had the greatest impact on the water quality status of the river, followed by mercury, arsenic, and selenium concentrations. This suggests that high mercury alluvial mining and extensive soil digging runoff entering the river associated with agricultural and animal grazing activities may account for the extremely poor quality status and unsuitable status of the River for any significant human activity based on its current quality status^5,22.

Pearson correlation

The Pearson correlation coefficient (r) was utilized to discover common origins and pathways influencing the distribution and interrelationship of variables in sediments, fish, and water ¹¹. The r-values closer to -1 and + 1 indicate strong negative and positive relationships, while 0 indicates no linear relationship among the variables. If their relationship is strongly positive, PTEs and pollutants may imply mutual dependence, behavioral similarities, or a common enrichment source ^3,11.

Table S1 shows the correlation coefficients for arsenic (As), mercury (Hg), selenium (Se), turbidity, electrical conductivity (EC), and pH for water, fish species, and sediment. A strong positive correlation was found between Hg and As in fishes, Se and pH in sediments, and Se and turbidity in water samples. The elevated Hg shows high As concentrations in fish species may be attributed to accumulation from similar sources of pollution such as alluvial mining. In sediments, the neutral pH medium indicates selenium’s stability and bioavailability even in growing concentrations. In water samples, Se concentrations may coagulate or adsorb onto dirty particles in the water affecting the impact and occurrence of Se in toxicity minimization of As and Hg. The positive strong correlation indicates similar sources of enrichment, transport, and accumulation while variables with no relationships indicated sources of enrichment from non-point or different sources of accumulation.

Principal component analysis

Principal component analysis (PCA) was implemented to reduce the number of variables as data sets of quality parameters with significant eigenvalues greater than one to two (2) significant principal components ³. Table S2 displays component loadings, proportional and cumulative percentages for datasets in fish, sediment and water samples. Strong factors (> 0.7), moderate (0.50–0.7), and weak factors (< 0.50) are the three categories of factor loadings that were used to explain the sources of pollution and transport in the various matrices.

The first principal component (PC1) showed 39.300% of the cumulative proportion in fish datasets. Strong and moderate loadings of Hg and As in PC1 indicates co-occurrence and accumulation of multiple elevated concentrations of in fishes which may be attributed common sources of pollution such as alluvial mining.

The first PCA accounted for 49.500% of the cumulative proportion in sediment data sets. Strong and moderate loadings were shown between the pH and Se, indicating that the transformation of Se species into more neutralizing forms are dynamic tools that affect Se occurrence, distribution, and bioavailability in the Offin River. The hydroxyl group may favor the adsorption of Se onto the sediment. The second PC2 accounted for 73.500% of the cumulative proportion. Hg showed moderate loading in PC2 indicating high sediments acting as a sink for accumulating residual and disposed of mercury from alluvial mining activities which may be converted into toxic forms.

In fishes, the first PCA accounted for 54.900% of the cumulative proportion. Strong and moderate loadings between Hg and As may indicate toxic forms of methylmercury and inorganic arsenic may be influenced by common transport processes such as alluvial mining. Weak and negative loadings in fish species, water, and sediments may be attributed to various contaminant processes and a lack of comprehensive pollution profile on all environmental matrices.

Average daily intake

The average daily intake exposure from consuming fish, water, and sediments contaminated with PTE from the Offin River is shown in Table S3. The exposure assessment aimed to establish that children or adults were more exposed through the ingestion pathway. The oral reference dose of each PTE under investigation was contrasted with the exposure assessment. The exposure order of PTEs based on their average values, which can be arranged as follows: In children, Hg > As > Se; in adults, As > Hg > Se. Also, the exposure to the environmental matrices was in the order of water, fish, and sediments. Exposure to the water and fish samples from Offin River via ingestion will make children and adults more prone to health issues than the sediment samples. These findings agree with previous research by showing that PTE exposure to sediments, water, and fish has a greater impact on children than on adults ^{3,6,11,12,35,38}. This observation may be because children have lighter bodies and engage more frequently in outdoor activities such as swimming in the river and crawling around the river bed than adults ^{3,6,11,12,35,38}. In comparison, the ADI values obtained were higher than the Rfd values (Table S3) for the various oral exposure routes in both adults and children.

Non-carcinogenic risks

For adults and children, estimates for the cumulative effects of non-carcinogenic risk (HI) in fishes, sediments, and water through the oral route are shown in Table 4. The HI of fish, water, and sediments were 5.240, 21.200, and 0.002 in children and 1.872, 9.130, and 0.001 in adults. The NCR findings indicate that HI values were greater than one (HI > 1) in all water and fish samples, suggesting a high likelihood of adverse health impacts related to the consumption of water and fish samples around the Offin River. These findings agree with previous studies conducted in Ghana near the Pra and Ankobra Rivers ⁶ and disagree with studies in Ghana near the Asuoyeboah River and Asafo market that fish and water samples posed less likelihood of detrimental health effects (HI < 1) ^12,39.

Table 4.

Non-carcinogenic and carcinogenic risk estimates via oral exposure of adults and children to PTEs in fish, sediments, and water from the River Offin.

Water	Variables	Mean	Min	Max	Standard	t-stats	t-critical	p-value	UCL	LCL
HQ	Hg	1.520	1.200	1.600	1	9.750	1.833	2.21E − 06	1.641	1.399
Children	Se	0.358	0.288	0.408	1	− 45.681	1.833	2.89E − 12	0.389	0.326
	As	19.320	15.600	22.800	1	23.843	1.833	9.59E − 10	21.058	17.582
	HI	21.198	17.512	24.784	1	25.521	1.833	5.24E − 10	22.988	19.407
HQ	Hg	2.100	0.571	2.714	1	5.773	1.833	1.34E − 04	2.531	1.669
Adults	Se	0.128	0.103	0.146	1	− 173.679	1.833	1.77E − 17	0.139	0.116
	As	6.900	5.571	8.143	1	21.500	1.833	2.40E − 09	7.521	6.279
	HI	9.127714	7.54	10.99429	1	22.692	1.833	1.49E − 09	9.937959	8.31747
Sediment
HQ	Hg	8.72E − 04	8.00E − 04	9.20E − 04	1	− 7.65E + 04	1.833	2.84394E − 41	9.02E − 04	8.42E − 04
Children	Se	4.46E − 05	4.08E − 05	4.80E − 05	1	− 1.12E + 06	1.833	8.99959E − 52	4.67E − 05	4.26E − 05
	As	1.37E − 03	1.20E − 03	1.56E − 03	1	− 2.45E + 04	1.833	8.06086E − 37	1.46E − 03	1.28E − 03
	HI	2.28E − 03	2.13E − 03	2.52E − 03	1	− 2.61E + 04	1.833	4.57155E − 37	2.37E − 03	2.20E − 03
HQ	Hg	3.11E − 04	2.86E − 04	3.29E − 04	1	− 2.14E + 05	1.833	2.67489E − 45	3.22E − 04	3.01E − 04
Adults	Se	1.59E − 05	1.46E − 05	1.71E − 05	1	− 3.14E + 06	1.833	8.50527E − 56	1.66E − 05	1.53E − 05
	As	4.89E − 04	4.29E − 04	5.57E − 04	1	− 6.86E + 04	1.833	7.55994E − 41	5.22E − 04	4.56E − 04
	HI	8.16E − 04	7.60E − 04	9.01E − 04	1	− 7.31E + 04	1.833112933	4.26473E − 41	8.47E − 04	7.85E − 04
Fish
HQ	Hg	0.207	0.165	0.257	1	− 40.562	2.132	1.10E − 06	0.261	0.152
Children	Se	0.001	0.0005	0.0010	1	− 9776.429	2.132	3.28E − 16	0.001	0.000
	As	5.035	2.772	6.769	1	5.636	2.132	2.44E − 03	7.022	3.047
	HI	5.242	2.937	7.020	1	5.823	2.132	2.17E − 03	7.265	3.219
HQ	Hg	0.074	0.059	0.092	1	− 132.589	2.132	9.70E− 09	0.093	0.054
Adults	Se	0.000	0.000	0.000	1	− 27,385.751	2.132	5.33E − 18	0.000	0.000
	As	1.798	0.990	2.418	1	3.122	2.132	1.77E − 02	2.508	1.088
	HI	1.872	1.049	2.507	1	3.352	2.132	1.43E − 02	2.595	1.150

Water	Dependents	Mean	Min	Max	Max RfD	t-stats	t-critical	p-value	UCL	LCL
CR	Children	2.42E − 02	1.95E − 02	2.85E − 02	1.00E − 04	25.040	1.833	6.21E − 10	2.63E − 02	2.20E − 02
	Adults	1.04E − 03	8.36E − 04	1.22E − 03	1.00E − 04	22.715	1.833	1.47E − 09	1.12E − 03	9.52E − 04
Sediment
CR	Children	2.05E − 07	1.80E − 07	2.34E − 07	1.00E − 04	− 16,309.475	1.833	3.12E − 35	2.19E − 07	1.91E − 07
	Adults	7.33E − 08	6.43E − 08	8.36E − 08	1.00E − 04	− 45,726.896	1.833	2.91E − 39	7.82E − 08	6.83E − 08
Fish
CR	Children	7.55E − 04	4.16E − 04	1.02E − 03	1.00E − 04	6.102	2.132	1.83E − 03	1.05E − 03	4.57E − 04
	Adults	2.70E − 04	1.49E − 04	3.63E − 04	1.00E − 04	4.425	2.132	5.73E − 03	3.76E − 04	1.63E − 04

Carcinogenic risk

The carcinogenic risks (CR) in children and adults from the ingestion of fish, water, and sediment in the River Offin are shown in Table 4. The mean CR values in children showed an estimated value of 7.55E − 04 in fish, 2.42E − 02 in water, and 2.05E − 07 in sediments, while a CR of 2.70E − 04 in fish, 1.04E − 03 in water, and 7.33E − 07 in sediments was obtained for adults. The estimated CR for arsenic in fish and water samples was greater than the maximum standard cancer risk tolerable intake value of 1E − 04 in children and adults. The findings of this study imply chronic adverse liver, lung, gastric, and skin cancer signs and symptoms in children and adults. The bioaccumulation of carcinogenic PTE such as arsenic in adults and children varies with size, age, food source, dietary intake, and exposure levels, therefore, this level of risk is guaranteed to increase. Based on the CR risks for adults and children, the agricultural, industrial and domestic use of water, fish and sediments from River Offin was not safe to use.

Bioconcentration factor

The Bioconcentration factor (BCF) estimates how fish in the River Offin accumulate the PTE present from the water body through their skin contact, gills, and other parts or tissues. The BCF will confirm that the water, which is the medium of transport of PTE present through ingestion by the fish, will be responsible for increasing the levels of PTE in the fish and causing toxicity to the fish and the consumer. The aquatic bioaccumulation models showed bioconcentration factors (BCF) of PTE in Table 8. The BCF in fishes ranges from 1.866 to 135.947 L/kg with a mean of 132.803 L/kg. Arsenic showed the highest BCF in fish, followed by Hg. The fish had a BCF of 135.947 L/kg for Hg, which means that the Hg concentration in the fish samples is one hundred and thirty-five times greater than in the Offin River water, showing significant bioconcentration and can induce a high level of toxicity to consumers. The fish had a BCF of 260.596 L/kg for As, which means that the As concentration in the fish samples is two hundred and sixty times greater than in the Offin River water, showing significant bioconcentration. The fish had a BCF of 1.866 L/kg for Se, which means that the As concentration in the fish samples is approximately two times greater than in the Offin River water, showing significant bioconcentration. The mean BCF of 132.803 L/kg was greater than 1, indicating that a considerable concentration of PTE in fish accumulates higher concentrations of As and Hg making Se synergistic and antagonistic effects less effective. Lower BCF values of less than 1 were reported in the Asuoyeboah River, Ghana¹².

Biota-sediment accumulation factor

The biota-sediment accumulation factor (BSAF) estimates how fishes accumulate the PTE present from the river bed (sediments) through their skin contact, gills, and tissues. The BSAF implies that the concentration of PTE in sediment influences the PTE concentrations detected in fish species exposed to or ingesting sediment particles and causing toxicity to the fish and the consumer. The biota-sediment accumulation factor (BSAF) is shown in Table S4. The BSAF in fish ranges from 0.015 (Se) to 3.680 (As), averaging 1.311. Arsenic showed the highest BSAF in fish, followed by Hg. A BCF of 3.8 for arsenic indicates that the concentration of As in fish samples is four times that of the sediment, indicating considerable bioaccumulation, a growing or accelerated absorption process, and possible ecological and health potential risks to humans as a consequence of PTE bioaccumulation. In comparison to our study, similar BSAF values of greater than 1 were reported in the Asuoyeboah River, Ghana¹².

Synergistic and antagonistic effects: molar ratios

The molar ratio of Se to Hg and Se to As can serve as a more accurate baseline index for assessing the health risk of river seafood based on using Se as an inhibitor to the toxic implications of mercury while the As: Hg molar ratio will serve as an indicator of the enhancement of As toxicity in fish and consumers. The molar ratios for Se: Hg, Se: As, and As: Hg have been provided in Table 5.

Table 5.

Selenium health risk–benefit values and molar ratios of selenium to mercury or arsenic and arsenic to mercury in fish species from River Offin, Ghana.

Variables	Se:Hg	Se:As	As:Hg	As-HBVSe	Hg-HBVSe	Total HBVSe
Mean	0.143	0.007	21.707	− 10.080	− 0.021	− 10.101
MIN	0.083	0.004	15.011	− 16.020	− 0.038	− 16.038
MAX	0.255	0.011	29.468	− 3.753	− 0.008	− 3.769
Standard	1.000	1.000	1.000	1.000	1.000	1.000
t-stats	− 29.445	− 733.669	8.200	− 4.290	− 202.026	− 4.293
t-critical	2.132	2.132	2.132	2.132	2.132	2.132
p-value	3.96E − 06	1.04E − 11	6.03E − 04	6.37E − 03	1.80E − 09	6.36E − 03
UCL	0.224	0.011	28.719	− 2.909	− 0.007	− 2.922
LCL	0.062	0.003	14.696	− 17.251	− 0.035	− 17.279

The molar ratio of Se: Hg showed a minimum value of 0.083 to 0.255 with an average value of 0.143. The mean Se: Hg molar ratio values recorded were < 1 indicating no free Se to protect against Hg toxicity. Hence, Hg poisoning and accumulation could occur through the consumption of Hg-contaminated fish. In comparison to our study, similar reports of a Se: Hg less than the protective value of 1 were reported in freshwater species, in Canada ¹⁸ while a Se: Hg molar ratio greater than 1 was reported in fish species in Oman sea, Iran ¹⁷, and Northeast Atlantic fish communities ⁴⁵.

The molar ratio of Se: As showed a minimum value of 0.004 to 0.011 with an average value of 0.007. The mean Se: As molar ratio values recorded were < 1 indicating no free Se to protect the fish and consumer against As toxicity. Hence, As poisoning and accumulation could occur through the consumption of As-contaminated fish. A similar observation of a Se: As molar ratio of less than 1 was reported in fish species in the uMgeni River, South Africa ⁴⁶.

The results showed that organic Se cannot effectively inhibit the accumulation of As and Hg in fish, reduce As and Hg bioaccessibility, and oppose the conversion of As and Hg from inorganic to organic form.

The molar ratio of As:Hg ranged from 15.011 to 29.468, with an average value of 21.707. The mean As: Hg molar ratio value recorded was > 1, signaling a synergistic effect of As on Hg and possibly increasing the toxicity of Hg in fish through consumption. In comparison to this study, similar observations of high arsenic to mercury ratio As:Hg greater than 1 have been reported in fish species from the Adriatic Sea in the Montenegrin Coast ⁸⁹, Oman Sea, Iran ¹⁷ and fish markets in Indonesia ⁹⁰.

The results from the molar ratios indicate that there is less Se to antagonize Hg and As in fish hence this may cause brain, heart, gills, and gastrointestinal disruption amongst other parts of the fish body as various types of PTEs, levels, and frequency may assert their influence on the fish species in the Offin River. Also based on the synergistic effect, As enhances Hg toxicity which in turn overrules the ability of Se to combat Se and As causing further disruption to the organs and tissues of the fish species.

Selenium health benefit value

The selenium health benefit value (Se-HBV) is a metric designed to evaluate the nutritional and beneficial advantages of selenium as well as its ability to protect against arsenic and mercury toxic consequences, specifically when it is associated with ingesting fish^17,46,47.

The cumulative and individual effects of Se-HBV evaluation on Hg and As toxicity sequestration are shown in Table 5.

The Se-HBV on Hg toxicity ranged between − 0.038 to − 0.008 with an average value of − 0.021. The Se-HBV on As toxicity ranged between − 16.020 to − 3.753 with a mean value of − 10.080. The total Se-HBV on As/Hg toxicity during the consumption of fish species from Offin River ranged between − 16.038 to − 3.769 with an average value of − 10.101.

The negative Se-HBV value in this study represents health problems related to Hg/As exposure, suggests that the selenium concentration in fish cannot mitigate the detrimental effects of arsenic/mercury or shows an unhealthy fish for human consumption. Furthermore, the Se-HBV’s magnitude reveals the deficiency (−ve) of Se associated with fish consumption.

In comparison to our study, similar observations of negative Se-HBV were reported in fish species from Atlantic fishing communities in Norway ⁴⁵ and herbivorous fishes from the Tarpajos River, Brazil ²⁴. However, positive Se-HBV values were reported in marine species such as aquatic species from coastal lagoons in Gulf California ⁴⁷ and commercial fish species from Turkish waters ¹⁴. The difference in magnitudes might be due to the varying influence of anthropogenic activities on the co-contamination of As/Hg species within the study area.

Pollution potential of PTE in sediments, fish, and water from the Offin River

The pollution potential of PTE in sediments, fish, and water from the Offin River is shown in Table S5.

The heavy metal evaluation index (HMEI) was used to assess the total risk level of certain PTE in a sample while considering any possible health hazards related to these PTE. The heavy metal evaluation index (HMEI) showed 9.222 in sediments, 6.900 in water samples, and 426.680 in fish samples. The HMEI values were less than 10, indicating low immediate health risk challenges and limited PTE bioavailability of PTEs despite a high contamination status.

Also, the contamination degree (Cdeg) of PTE showed 6.222 in sediments, 3.900 in water samples, and 423.680 in fish samples. The Cdeg value greater than 3 indicates severe and alarmingly high contamination status across multiple PTEs in fish, water, and sediments. The high Cdeg indicates potential risks to human health associated with the accidental ingestion of sediment or consumption of fish and water. The Cdeg and HMEI show a higher contamination status in fish than in the river and sediment samples.

The Cdeg and HMEI in this study agree with the BCF and BSAF, showing that the PTE toxicity is increasing in the fish samples more than in the sediment and river water environment of the various species analyzed in the Offin River.

Mitigation strategies and policy implementations

Strong environmental awareness in conjunction with environmental protection agencies on water resources and the environment should be consistently disseminated to inhabitants around the country via media interventions, and other educational programs to sensitize the youth about environmental degradation. Given this report, Ghana—Water Resource Commission (WRC), Environmental Protection Agency (EPA), National Commission for Civic Education (NCCE), and Food and Drugs Authority (FDA) should proactively enforce stricter working management and protection laws, monitoring programs, environmental awareness programs, and treatment interventions to control and mitigate the contamination upsurge in River Offin devoid of political interference.

Non-governmental organizations (NGOs), WRC, EPA, NCCE, local and residential communities, and industries within mining communities should work together to manage the sources of severe contamination and ensure sustainable protection of water resources. They should also set regulations, enforce compliance, and continuously monitor and evaluate the pollution crisis in the affected region to lessen environmental degradation. These will sustain all ecosystems that rely on clean water and aquatic life. In addition, future research needs to consider long-term monitoring programs, environmental awareness, sophisticated and economical remediation techniques that need to be explored, and intensive investigations of the chronic exposure effects using relative biomarkers are required to properly ascertain the immediate health challenges to populations within the community. However, managing the quality pollution status via broad strategies and policies and citizen education is crucial to ensuring long-term sustainability and operability for aquatic life and human safety.

Limitations and recommendations

The study may have some limitations, as the samples were from the Offin River only. Greater and larger sampling and analysis across all regional and local rivers and water bodies will enrich the findings and enhance policy implementations.

Across all rivers and aquatic habitats, efforts should be made to monitor and establish an in-silico, in-vivo, and in vitro assessment of Se effects on Hg and As in rats, which will create a baseline study for treatment studies and policy implementations nationwide.

In Ghana, and most African countries, there has been little attempt to establish the antagonistic effects of Se, Hg, and As since the focus has been on the toxicity of toxic elements but emerging contaminants and their health risks should be discussed to further iron out and add to the existing knowledge about Se interplay on other PTEs.

Due to financial constraints, future studies should consider the seasonal variations of the environmental matrices were not considered to further ascertain the season that contributes more destruction and occurrence of contaminants in the environment.

For other PTEs, contaminants should be screened in soil, air, plant, and other environmental species to establish the elemental interplay amongst other PTEs, establish a comprehensive toxicological profile of the status of the environmental community, and further find ways of solving them. Furthermore, an economical treatment technology should be implemented to treat the deteriorated nature of the brownish water body and preserve aquatic life.

Conclusions

This study has shown significant impact on the physicochemical status of River Offin with significant concentrations of As and Hg in water, sediment and fish. This study also showed a poor water quality condition of the river with turbidity being the main driving force of polluting the aesthetics and physical attributes of the Rover Offin in Ghana. Elemental interactions ratios, pollution status of PTE occurrence, and health risk were also assessed for children and adults revealing significant implications to local communities.

The study found that the average concentrations of Hg, As, EC, Se, and turbidity exceeded recommended limits in all water samples, while mean Hg and Se also surpassed threshold limits in sediment samples. Additionally, mean Hg and As levels were higher than standard limits in fish species, increasing toxicity challenges in aquatic ecosystems (River Offin).

The water quality index indicates the River’s quality is extremely poor and unsuitable for drinking, household, agricultural, or industrial activities.

The BSAF and BCF models indicated that the accumulation of Hg and As in fish was 132 times greater from water than from sediment, highlighting the River as the primary source of potentially toxic elements (PTE) and contaminant transport.

The analysis of molar ratios reveals that selenium (Se) did not mitigate the toxicity of mercury (Hg) or arsenic (As). Conversely, the As:Hg ratio greater than 1 indicates that arsenic enhances mercury toxicity in fish, posing significant risks to aquatic life. The accumulation of potentially toxic elements (PTEs) can lead to severe reproductive deformities, mortality, and degradation of aquatic habitats in the river ecosystem.

The average daily intake (ADI) shows that the exposure risk assessment indicated that children will suffer from exposure to the deleterious effects of PTE than adults. The cumulative non-carcinogenic risk associated with the oral consumption of water and fish was greater than one (HI > 1), indicating kidney, liver, diarrhea, stomach upset, memory loss, and reproductive disorder in both children and adults. The carcinogenic risk (TCR) associated with As oral exposure to fish and water was less than the tolerable intake value of 1.00E − 4, exposing children and adults to cancer of the brain, respiratory system, lungs, kidney, colon, and skin.

The contamination degree and heavy metal pollution evaluation index (HMEI) indicated a severely high contamination status for multiple PTEs (Hg, As, Se) in fish, water, and sediment samples. According to pollution potential studies, although there is a high pollution potential from PTEs occurrence in water and fish samples, there is a strong need for cleanup and water restoration interventions.

The principal component analysis and Pearson correlation analysis indicated turbidity, EC, Hg, and As occurrence in River Offin may be attributed to mercury-based gold mining practices, while the absence of Se-rich rocks and low salinity by accompanied by high turbid water bodies.

The acknowledgement of primary limitations such as a smaller sample size and program, funding constraints, and fewer water quality parameters analysed, which may have impacted the quality and in-depth nature of the water quality assessment results.

The necessity for robust environmental monitoring and management in the River Offin due to high PTE concentrations. It advocates for regular surveillance, strict enforcement to mitigate degradation, and protection for local populations reliant on the river’s water. The findings underline the urgent need for improved environmental policies to prevent further ecosystem harm and suggest raising awareness of health risks from contaminated fish and water, alongside proposing educational outreach for safe consumption practices.

Future research should explore the impacts of illegal mining and human activities on tributary rivers, assess seasonal variations, and evaluate biomarkers in urine and blood across various age groups. Additionally, river restoration should be considered as a significant factor in mining and environmental regulation.

Author contributions

Gerheart Winfred Ashong: Conceptualization, Data curation, Formal analysis, Methodology, Supervision, Validation, Visualization, Writing—original draft, and Writing—review & editing Boansi Adu Ababio: Conceptualization, Data curation, Formal analysis, Resources, Validation, Visualization, Writing—review & editing Edward Ebow Kwaansa-Ansah: Conceptualization, Methodology, Project administration, Resources, Supervision, Writing—review & editing Joseph Appiah-Kubi: Investigation, Methodology, Resources.

Data availability

The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Declarations

Competing interests

The authors declare no competing interests.