Risk assessment of polychlorinated biphenyls (PCBs) in surface and groundwater in three states in Nigeria Open Access

Polychlorinated biphenyls (PCBs) are globally recognized as major environmental pollutants due to their persistence, bioaccumulative nature, and toxic effects on both humans and wildlife. These synthetic organic chemicals were extensively produced and used in industrial applications for their excellent dielectric properties, hydrophobicity, thermal stability, and resistance to degradation (Lauby-Secretan et al. 2018). PCBs have historically been used in products such as electrical transformers, capacitors, paints, plastics, adhesives, and hydraulic fluids, contributing to their widespread adoption (Wimmerová et al. 2020). However, concerns over their environmental persistence and long-range transport, coupled with evidence of their harmful health effects, led to a global ban under the Stockholm Convention on Persistent Organic Pollutants (UNEP 2018).

Despite this ban, PCBs remain a significant environmental concern, particularly in developing countries where these compounds are still present in old equipment, waste sites, and contaminated environments (Iwegbue 2020). Numerous studies have demonstrated the adverse effects of PCB exposure, including endocrine disruption, immunotoxicity, and carcinogenicity in humans, as well as reproductive failures in wildlife (Ediagbonya et al. 2015; Wang et al. 2018; Klaren et al. 2019; Haddaoui et al. 2021). PCB exposure has also been linked to deoxyribonucleic acid (DNA) sequence changes and oxidative stress, which may exacerbate health risks (Zhang et al. 2022). Recent studies have also highlighted the contamination of water sources by both heavy metals and persistent organic pollutants such as PCBs, emphasizing the risks associated with these pollutants in rainwater harvesting systems and rivers (Ediagbonya & Gbolahan 2018; Ediagbonya et al. 2022; Halfadji et al. 2022).

Water pollution, especially with persistent organic pollutants like PCBs, is a pressing environmental issue. Rivers and wells are critical sources of drinking water, agricultural irrigation, and industrial use, yet they are increasingly contaminated by runoff, industrial discharge, and improper waste management (Igwe et al. 2018). In Nigeria, limited studies have focused on the contamination of rivers and groundwater by PCBs, leaving a significant research gap (Ediagbonya et al. 2022, 2023a, 2023b, 2023c). Understanding the contamination levels in these water bodies is essential for developing effective mitigation strategies and ensuring public health safety.

The Forcados River, located in Bomadi LGA, is a tributary of the Niger River. This river spans an average width of 1.6 km and supports key ecosystem services such as fisheries, agriculture, and transportation for the local communities. The surrounding region is characterized by tropical rainforest vegetation, with average annual rainfall exceeding 2,500 mm. The area experiences a humid tropical climate, with temperatures ranging between 23 and 33 °C throughout the year. Economic activities along the river include fishing, small-scale farming, and jetty operations. However, the river faces significant pollution from oil spills, wastewater runoff, and improper waste disposal (Igwe et al. 2018).

Lake Eleyele, located northwest of Ibadan in Ido LGA, Oyo State, was created in 1939 by damming the Ona River. The lake has an area of approximately 1.1 km² and is fed by several streams, including Otaru, Awba, Yemoja, and Alapo. It provides ecosystem services such as irrigation, flood control, and water supply for domestic and industrial purposes. The region receives average rainfall of about 1,200 mm annually, with a peak during the months of July and September. Temperatures in the area range from 24 to 32 °C. Economic activities include agriculture, fishing, and sand mining. Pollution sources in the lake include agricultural runoff, urban wastewater, and siltation from nearby construction activities (Imevbore 2019).

The wells in Apete, Ido LGA, are shallow groundwater sources used by local residents for drinking and domestic purposes. These wells are located near residential areas, often close to abandoned ponds and poorly managed waste sites. The surrounding area experiences an average annual rainfall of 1,300 mm, with temperatures ranging between 23 and 34 °C. Economic activities include small-scale farming, trading, and informal waste management. Due to inadequate depth and proximity to contamination sources, these wells are highly susceptible to pollution from heavy metals, microorganisms, and chemical spills (Igwe et al. 2018).

The selected sites represent critical water sources for their respective communities. The Forcados River and Lake Eleyele provide fishing and irrigation opportunities, while the wells in Apete serve as a primary drinking water source. Understanding the extent of PCB contamination in these locations is vital for assessing health risks and developing mitigation strategies. These sites also reflect typical water resource challenges in developing countries, where industrial and agricultural activities threaten ecosystem integrity.

The water samples were collected from three selected sites: the Forcados River (Bomadi LGA, Delta State), Lake Eleyele (Ido LGA, Oyo State), and Apete Wells (Ido LGA, Oyo State). These sites were chosen due to their ecological and hydrological significance, their proximity to potential pollution sources, and their importance to local communities. The Forcados River serves as a major tributary of the Niger River and is exposed to industrial and agricultural runoff (Igwe et al. 2018). Lake Eleyele is an essential reservoir for irrigation and domestic water use (Imevbore 2019), while Apete Wells are critical for drinking water in the region, often located near waste sites and abandoned ponds.

The selection of sampling sites was based on land-use patterns, proximity to pollution sources, and the hydrological significance of the water bodies. The Forcados River and Lake Eleyele were selected due to their susceptibility to industrial and agricultural pollutants, while the Apete Wells were chosen to assess contamination in groundwater from nearby human activities (Igwe et al. 2018).

A total of 40 surface water samples and 20 groundwater samples were collected to ensure spatial coverage of the study sites. Sampling points were determined based on accessibility and coverage, with a sampling frequency of approximately 5 km intervals along the rivers and lakes. Samples were collected during a single dry season to minimize hydrological variability and ensure consistency in results.

For each location, 1.5 L of water was collected in pre-cleaned glass sampling vials. Surface water samples were taken from the midstream of rivers and lakes at approximately 1 m depth using a weighted sampling device, while groundwater samples were collected from wells using a sterilized hand pump. The samples were labeled, sealed, and immediately transported to the laboratory under chilled conditions (4 °C) to prevent degradation. Table 1 provides detailed coordinates and the methods used for sample collection.

To ensure the reliability of results, method blanks, duplicates, and standard reference materials (SRM) were included. Samples were spiked with 13C12-labeled PCB congeners to validate recovery rates, which ranged from 89.9 to 99.8%. Data analysis was performed using the Statistical Package for the Social Sciences (SPSS) software (version 26), with all computations conducted at a 95% confidence interval. The calibration lines for PCB concentrations demonstrated r² values ranging from 0.9995 to 0.9999, ensuring high precision in quantification (SERAS 2006).

To ensure the reliability and accuracy of the data, multiple quality control and assurance measures were implemented. Method blanks, surrogate standards, and duplicates of the SRM were used to evaluate the representativeness, precision, and repeatability of the analyses. Matrix-spiked samples and 13C12-labeled PCB congeners were employed to validate the recovery of targeted compounds. Recovery rates for the 13C12 PCB congeners ranged between 89.9 and 99.8%, while PCB spike recovery rates ranged from 69.9 to 98.5%. These recovery rates fall within acceptable ranges established in similar studies (e.g., 70–120% for environmental monitoring; Reddy et al. 2019) and highlight the effectiveness of the analytical procedures used.

Analytical precision was maintained by calibrating the gas chromatography-mass spectrometry (GC-MS) system using five-point serial dilution standards. Limits of detection (LOD) and quantification (LOQ) were calculated based on the signal-to-noise ratios of 3:1 and 10:1, respectively. For PCB congeners, LOD values ranged from 0.3 to 1.7 ng/g, while LOQ values ranged from 1.0 to 5.0 ng/g. These values are significant in environmental monitoring as they ensure the detection of PCB concentrations even at trace levels, allowing for robust assessment of contamination risks (Zhao et al. 2019).

All data analyses were conducted using the SPSS, version 26. Analytical procedures included external calibration and the use of r² values (0.9995–0.9999) to ensure accuracy in quantification. The mean PCB levels for each location were compared using analysis of variance (ANOVA) at a 95% confidence level (p = 0.05). These statistical methods facilitated the identification of significant differences in PCB concentrations across the study sites.

By adhering to rigorous quality control measures and statistical analysis, the study ensures the reliability and validity of its findings, providing a strong foundation for further environmental assessments.

The extraction of PCBs from surface and groundwater samples followed the procedures outlined in SERAS (2006). Methylene chloride was used as the extraction solvent due to its high efficiency in separating hydrophobic organic compounds like PCBs from aqueous matrices. The process was carried out at a neutral pH, achieved by adjusting with a 1:1 sulfuric acid solution or 10 N sodium hydroxide. Maintaining a neutral pH is critical for preventing the degradation or loss of PCB congeners during the extraction process (Reddy et al. 2019). This approach ensured the preservation of analyte integrity and improved recovery rates.

Approximately 1 L of water from each sample was extracted in series with methylene chloride. Process blanks and laboratory control samples were prepared by adding surrogate working solutions of 200 ng/mL to the samples before extraction. The extracts were concentrated to a final volume of 10 mL to enhance the detectability of PCBs. Excess pressure was vented periodically during the process, and the separating funnel was shaken for 2 minutes to ensure efficient mixing.

For solvent exchange, 60 mL of hexane was added to the extracts, which were then filtered through a funnel containing glass wool and anhydrous sodium sulfate to remove moisture and particulates. The combined extract was concentrated to a final volume of 1 mL and analyzed using GC-MS.

AccuStandard supplied PCBs standard, 2,000 ppm (Catalogue Number: M-8080) with 17 PCB components. The GC-MS was calibrated using five-point serial dilution calibration standards (1.00, 0.25, and 0.05 ppm) taken from stock. Prior to calibration, the abundance of m/z 69, 219, and 502, as well as other instrument optimum and sensitivity parameters, were validated using stated criteria to auto-tune the MS to perfluorotributylamine. To accomplish low-level detection of the target components, the amount of PCBs in the sample was measured by GC-MS in selective ion monitoring (SIM) and scan mode. The Agilent Technologies 5975C inert mass spectrometer was coupled to an Agilent 7820A gas chromatograph. It had a triple-axis detector and an electron-impact source. Agilent Technologies' HP-5 capillary column (30 m length × 0.32 mm diameter × 0.25 m film thickness) coated with 5% phenyl methyl siloxane served as the stationary phase for compound separation. Helium was used as the carrier gas, with an average velocity of 40.00 cm/s and an initial nominal pressure of 026 psi. The gas was used at a constant flow rate of 1.2 mL/min. In splitless mode, the samples were injected into a 1 mL volume at 250 °C. When the gas saver mode was turned off, the purge flow to the overflowing vent was 30.0 mL/min for 0.35 min, and 31.24 mL/min overall. After 1 minute at 110 °C, the oven was immediately heated to 310 °C for 3 min, with a 15 °C increase per minute. With a 3-min solvent delay, the run took 16 min. The mass spectrometer was configured for electron-impact ionization at 70 eV, with the quadrupole temperature at 150 °C, the transfer line temperature at 300 °C, and the ion source temperature at 230 °C. To obtain the ions, two modes of operation were employed: selective ion mode (SIM) and scan mode, which involved scanning from m/z 35 to 550 amu at a rate of 2.0 s per scan. Following calibration, the samples were analyzed to determine the appropriate PCB content.

Long-term daily doses for oral, respiratory, and skin contact are denoted by CDIing-nc, CDIinh-nc, and CDIderm-nc, respectively. The PCB concentrations in water samples that meet the UCL95% are referred to as CUCL (mg kg⁻¹). In this case, people were supposed to ingest 100 mg of water each day, or IngR. An annual exposure frequency of 350 days was examined, excluding 15 vacation days. The exposure duration is stated in years. ‘ABSD’ stands for ‘dermal absorption factor’ (0.14 for PCB), ‘SA’ for ‘surface area of the skin in contact with water’ (cm⁻²), ‘AF’ for ‘skin adherence factor for water’ (mg cm⁻²), ‘ATnc’ for ‘averaging time for non-carcinogens’ in days, and ‘CF’ for ‘conversion factor’ (1 × 10⁻⁶ kg mg⁻¹).

Table 2 shows the mean and standard deviations of PCB concentrations, as well as the ANOVA results and post-hoc tests represented by alphabetic superscripts. The ANOVA shows significant differences in PCB concentrations among four locations (F-statistic = 2,434.310, p < 0.001). To discover which groups differ significantly from each other. Superscripts represent the outcomes of post-hoc tests that were carried out. In RIVER A1, di-PCBs (0.07 ± 0.01 mg/L) and hepta-PCBs (0.07 ± 0.01 mg/L) had similar mean values but differ statistically from penta-PCBs (0.65 ± 0.01 mg/L) and tri-PCBs (0.75 ± 0.04 mg/L). RIVER 3 has significantly higher mean concentration of total PCBs (32.04 ± 0.35 mg/L) compared to RIVER H₂O (13.61 ± 0.31ᵃ mg/L) and 10 cm WELL (22.59 ± 0.00 mg/L). Post-hoc tests provide a clear distinction between locations, allowing us to better understand PCB distribution and relevance in each setting. However, the United States prohibited the production of PCBs after it was determined that they were harmful to the environment and should be classified as persistent organic pollutants (Batang et al. 2016; Carrizo et al. 2021). Regardless of their mix, they are viscous liquids or solids that range in color from clear to yellow, have no taste or odor, are extremely stable, and can withstand high pressures and temperatures (Reddy et al. 2019). In this study, three river water samples and one well water sample were collected, and 17 PCB congeners were discovered. The findings demonstrated the dispersion and bioaccumulation of PCBs in various natural environments. In general, PCB levels were higher in river water than in groundwater. PCB values in river water vary from 0.01 ± 0.00 to 1.66 ± 0.01 mg/L. PCB 61 and PCB 5 have the highest mean concentrations, whilst 206 has the lowest. Unintentional spills and leaks during transit, as well as fires or product leaks containing PCBs, all contribute to PCB contamination of rivers (Gorshkov et al. 2017). River 3 has the highest levels of PCBs (32.04 ± 0.35 mg/L) due to jetty boat emissions, sewage dumps, and wastewater runoff from polluted areas (Net et al. 2015). River 3 and River H₂O have greater PCB5 concentrations than other test sites, which could be linked to flooding and pollution, both of which endanger persons and ecosystems (Rudel et al. 2008; Cohn et al. 2012). In contrast, 10 cm wells had less PCB congeners found than river samples, with PCB 47, 151, 161, 159, 183, and 189 all falling below detection limits (BDL). PCB 17 has the highest mean concentration among all samples (13.18 ± 0.05 mg/L), indicating a hazard to the environment. This is owing to the inadequate depth of the well, which draws water from an abandoned pond contaminated with heavy metals, microbes, leaks and spills. PCBs 5, 21, 61, and 187 were found to be significantly elevated in River A1. This is most likely due to point sources causing localized high PCB congeners. A range of other variables, such as hydrodynamics and crystal size, can also affect PCB concentration (Moraleda-Cibrián et al. 2015; Gao et al. 2018). With the exception of PCB 5 and 99, which have greater values in H₂O PCB, all other PCB congeners are below the EPA limit. This could be linked to the PCBs that these small-scale enterprises release into the river system. In May 2015, many stations identified a high level of five PCBs in the upper water layer (5 m) in the South Atlantic Basin, with a value of 3.1 ng/L. According to Unyimadu et al. (2018), Nigeria has no regulatory control limits for PCBs in water. To assess the importance of PCBs in these samples, we used the recommended PCB levels in water provided by a few international regulatory authorities. PCB levels are specified at 1,000 μg kg⁻¹ for the Dutch action value and the Australian and New Zealand Ecological Investigation Level (VROM 1994), 1,300 μg kg⁻¹ for the Canadian authority's water quality guidelines (CCME 2007), and 220 μg kg⁻¹ for the US EPA's health-based screening level for total PCBs, which aims to prevent harmful health effects associated with chronic exposure (Oksanen et al. 2019; Kustova et al. 2021).

Table 3 displays significant changes in PCB concentrations and their statistical significance, as determined by ANOVA and post-hoc tests, with alphabetic superscripts.

The study provides solid evidence for the dispersion of PCBs with varying chlorine atom counts. Di-PCBs, which contain two chlorine atoms, exhibit large concentration differences among locations. River 3 (20.69 ± 0.28) has significantly higher di-PCB levels compared to River A1 (8.09 ± 0.09^a), River H₂O PCB (10.59 ± 0.28), and 10 cm well (16.57 ± aa). Tri-PCBs, which contain three chlorine atoms, had very stable quantities across all locations, with no statistically significant alterations seen. The mean results (0.15 ± 0.05, 0.20 ± 0.01, 0.17 ± 0.01, and 0.14) are comparable across River A1, River 3, River H₂O PCB, and 10 cm well. Penta-PCBs, which contain five chlorine atoms, show remarkable variability. River 3 (10.98 ± 0.02^a) had significantly higher penta-PCB levels than River A1 (0.96 ± 0.01^a), River H₂O PCB (2.61 ± 0.03^a), and 10 cm well (4.56). Hepta-PCBs, which contain seven chlorine atoms, also differ dramatically among locations. River A1 (1.15 ± 0.03^b) and 10 cm well (1.26 ± aa) had higher mean hepta-PCB concentrations compared to River 3 (0.15 ± 0.02^b) and River H₂O PCB (0.17 ± 0.01^b). The ANOVA results demonstrate significant differences (F-statistic = 1,172.378, p < 0.001) among PCB groups in terms of chlorine atom count. Post-hoc studies can provide valuable information about the specific sites and PCB groupings generating these variations. The numbers demonstrate the difficulties of PCB distribution in different situations, emphasizing the importance of focused monitoring and management approaches to address potential environmental and health hazards associated with these persistent pollutants. PCB congeners in River 3 (Bomadi River) are largely di- and penta-chlorinated biphenyl, similar to the Niger Delta River (Iwegbue 2020). Also, 10 cm well samples show an increase in di- and penta-PCB. The increasing PCB burden in River 3 is due to pollution similar to that seen in the Udu River (Paschal et al. 2022). The degree of chlorination affects PCB metabolism in the human body. In general, less chlorinated PCBs metabolise faster than highly chlorinated ones (Lemieux 2019). The study's findings, which linked di-CBs to a higher proportion of PCB profiles, indicate that PCBs can remain in the environment despite pollution and impaired metabolism. PCB levels in surface water in Oyo and Delta are the result of runoff and long-term pollution from the garbage dump. In this environment, waste can only travel long distances, which may explain why some rivers have lower amounts of chlorinated PCBs. These compounds are also more likely to collect in the atmosphere and be delivered downward by precipitation (Gao et al. 2017).

The PCB concentrations recorded in this study (Forcados River, Lake Eleyele, Apete River, and Apete Wells) were compared with other studies conducted globally to contextualize the findings and identify potential patterns in contamination. Table 4 highlights the concentrations of PCB congeners across different water systems, reflecting varying levels of contamination influenced by industrial, agricultural, and urban activities.

In this study, the Forcados River exhibited the highest PCB concentration (32.04 ± 0.35 mg/L) compared to Lake Eleyele (10.38 ± 0.11 mg/L), Apete River (13.61 ± 0.31 mg/L), and Apete Wells (22.59 mg/L). The elevated levels in the Forcados River are consistent with the river's exposure to industrial discharges, jetty boat emissions, and urban wastewater, all of which contribute to PCB contamination. This aligns with findings by Paschal et al. (2022) in the Udu River, Nigeria, where concentrations ranged from 5.34 to 16.1 mg/L, further emphasizing that rivers near industrial or urban centers tend to accumulate higher PCB levels.

In comparison, Lake Eleyele's PCB concentration, though moderate, reflects significant contamination from agricultural runoff and nearby urban wastewater. This is consistent with studies in Egypt's River Nile, where PCB levels ranged from 14 to 20 mg/L (Ayman et al. 2015). The PCB contamination in lakes is typically influenced by runoff from surrounding land uses, which can carry pesticides and industrial residues. Similarly, South Africa's Umgeni River recorded concentrations ranging from 6.91 to 252.30 mg/L (Gakuba et al. 2015), demonstrating how varying hydrological and land-use factors shape PCB pollution in African water systems.

Groundwater sources, such as Apete Wells, displayed lower PCB concentrations compared to surface water systems. However, the 22.59 mg/L detected in Apete Wells is concerning, as it suggests contamination from nearby waste disposal sites and abandoned ponds. This pattern mirrors findings from Port Harcourt, Nigeria, where PCB concentrations ranged from 0.053 to 0.2733 mg/L (Njoku et al. 2016). Groundwater systems are less directly exposed to surface pollutants but can still be contaminated through infiltration of leachates or poorly managed waste.

The PCB concentrations in Nigerian rivers and lakes are comparable to those reported in industrialized countries. For instance, River Missouri in the USA recorded PCB levels between 11 and 250 mg/L (Kathy et al. 2018), while France's Ogre River had concentrations ranging from 120 to 170 mg/L (Khawla et al. 2012). These results highlight the pervasive nature of PCB pollution globally, with contamination driven by similar sources, such as industrial discharge and improper waste management.

Interestingly, the concentrations in Izmit Bay, Germany (5.67–14.81 mg/L, Leyla et al. 2012), and the Ebro River, Spain (2.9–37 mg/L, Alonso et al. 2011), are closer to those observed in Nigerian rivers and lakes, suggesting comparable contamination risks in regions with mixed industrial and agricultural activities.

The findings from Table 4 emphasize the global scale of PCB contamination and its significant implications for environmental health. Rivers and lakes, as primary water sources, are highly vulnerable to pollutants due to their direct exposure to runoff, industrial discharge, and atmospheric deposition. The elevated PCB concentrations in Nigerian water systems pose serious risks to aquatic ecosystems and human health, as PCBs are known to bioaccumulate in fish and other aquatic organisms, entering the food chain and potentially causing adverse health effects, such as endocrine disruption and cancer (Van den Berg et al. 2006).

The correlation matrix presented in Table 5 highlights the relationships between different PCB congeners across the study locations. Understanding these relationships is critical as it provides insights into the sources of PCBs, their environmental behaviors, and interactions within the aquatic systems. Positive correlations between congeners often indicate shared sources or similar environmental transport mechanisms, while negative correlations suggest different sources or opposing environmental behaviors.

The analysis revealed both strong positive and negative correlations between PCB congeners, reflecting the complexity of their distribution and interactions.

• 1. Positive correlations:

• PCB congeners such as PCB3 and PCB21 displayed a strong positive correlation (r = 0.812, p < 0.001), indicating a likely shared source, such as industrial emissions or wastewater runoff. This pattern aligns with findings from similar studies, where closely related PCB congeners were shown to originate from specific industrial processes, such as the use of electrical transformers and capacitors (Van den Berg et al. 2006).

Another notable positive correlation was observed between PCB151 and PCB161 (r = 0.965, p < 0.001). This strong relationship suggests that these congeners may have been introduced into the environment through a common pathway, such as agricultural runoff carrying PCB residues from pesticides. Such correlations are critical for identifying pollution sources and prioritizing mitigation strategies.

• 2. Negative correlations

• Conversely, strong negative correlations were observed, such as between PCB98 and PCB111 (r = −0.947, p < 0.001). These negative correlations may reflect differences in the sources of these congeners or variations in their environmental transport and degradation. For instance, PCB congeners with higher chlorine content, like PCB111, are more hydrophobic and tend to adsorb onto sediments, whereas less chlorinated congeners, such as PCB98, may remain in the water column (Zhao et al. 2019). This distinction highlights how environmental conditions, such as sediment composition and hydrology, influence PCB behavior.

• 3. Complex interactions

• Certain congeners, such as PCB183, exhibited moderate positive correlations with multiple other congeners, including PCB168 (r = 0.659, p < 0.001) and PCB206 (r = 0.666, p < 0.001). This suggests overlapping but not identical sources, potentially combining industrial runoff with atmospheric deposition. The moderate correlations may also reflect environmental processes, such as partial degradation or differential transport of PCB congeners in aquatic systems.

The observed correlations have significant implications for environmental health and risk management. Strong positive correlations, like those between PCB3 and PCB21, indicate a need to identify and mitigate specific point sources of pollution, such as factories or urban wastewater systems. On the other hand, negative correlations, such as between PCB98 and PCB111, suggest complex environmental behaviors that require further investigation, particularly in sediment-water interactions.

These findings also point to the potential for bioaccumulation and biomagnification in aquatic ecosystems. Congeners with strong positive correlations are more likely to co-occur in fish and other organisms, leading to cumulative exposure risks for both wildlife and humans. Studies have shown that certain PCB congeners are highly toxic and can cause endocrine disruption, immune suppression, and cancer in humans (Reddy et al. 2019). Addressing these risks requires a holistic approach, including source control, monitoring, and public awareness.

Table 6 highlights the commonalities of PCB congeners before and after extraction using principal component analysis (PCA). Communalities measure how much of the variance in each variable (PCB congener) is explained by the extracted components. The extraction values show the proportion of variance retained in each variable after dimensionality reduction, which is critical for understanding how well the PCA captures the essential information from the dataset (Gorshkov et al. 2017; Zhao et al. 2019).

The initial communalities for all PCB congeners were set at 1.000, indicating that 100% of the variance in each variable was initially retained. After PCA extraction, the commonalities decreased for most congeners, reflecting that only a portion of the original variance was explained by the principal components. For example, PCB5 showed a reduced commonality of 0.703, indicating that 70.3% of its variance was retained after extraction, while the remaining 29.7% was likely explained by other, less significant components.

Interestingly, some congeners retained nearly all of their variance post-extraction. PCB98 and PCB99 displayed unusually high communalities of 0.999 and 0.996, respectively, suggesting that these variables are highly relevant to the principal components and contribute significantly to the overall variance in the dataset. Similar findings have been reported in studies on PCB distributions in aquatic environments, where specific congeners demonstrated higher significance due to their persistence and environmental mobility (Moraleda-Cibrián et al. 2015; Gao et al. 2018). This finding indicates that PCB98 and PCB99 are pivotal in explaining the patterns in PCB distribution and may serve as key indicators of contamination sources or environmental processes.

The variation in communalities highlights the differential importance of PCB congeners in contributing to the overall structure of the dataset. Congeners with higher post-extraction commonalities, such as PCB111 (0.984) and PCB161 (0.992), are likely more strongly influenced by shared environmental sources or processes (Gao et al. 2017). Conversely, congeners with lower communalities, such as PCB159 (0.887) and PCB168 (0.827), may exhibit unique environmental behaviors or originate from distinct sources. These results are consistent with prior research emphasizing the role of environmental and physicochemical properties in shaping PCB distribution (Vorkamp 2018; Paschal et al. 2022).

The results suggest that PCA effectively identifies the most significant PCB congeners, enabling dimensionality reduction while retaining meaningful variance. This reinforces PCA's utility as a tool for simplifying complex datasets and focusing on key variables (Zhao et al. 2019).

Table 7 summarizes the total variance explained by each principal component extracted through PCA. This table is vital for understanding how much of the dataset's variability is retained by the principal components, helping to identify the most significant dimensions that capture the essence of the data (Gorshkov et al. 2017; Zhao et al. 2019).

The first principal component explains 38.484% of the total variance, and the second component contributes an additional 35.201%, bringing the cumulative variance explained by the first two components to 73.685%. These two components effectively summarize the majority of the dataset's variability, capturing the key patterns in PCB distribution.

The third component accounts for another 20.255% of the variance, resulting in a cumulative variance of 93.941% explained by the top three components. This high percentage underscores the efficiency of PCA in reducing the dataset's dimensionality while retaining nearly all of its original variability. Similar trends have been observed in environmental studies, where the first few components often capture the majority of the variance due to dominant contamination patterns (Moraleda-Cibrián et al. 2015; Gao et al. 2018). The remaining components (4 through 15) contribute minimal variance, with eigenvalues close to or below zero, making them statistically insignificant and less relevant for further analysis.

The dominance of the first three components suggests that PCB contamination patterns in the study are primarily influenced by a few key factors, such as shared pollution sources or environmental behaviors of specific congeners. For instance, industrial discharge, urban runoff, and sediment transport dynamics are likely the primary drivers of variability in PCB distribution, while other minor factors play less significant roles (Vorkamp 2018; Paschal et al. 2022).

The cumulative variance explained by the top three components highlights PCA's utility as a data-reduction tool. This allows researchers to focus on the most significant dimensions of PCB contamination, simplifying complex datasets into actionable insights. Such dimensionality reduction is particularly valuable in environmental studies where datasets often contain numerous variables with overlapping or redundant information.

The high percentage of variance explained by the first three components enables researchers to prioritize the most critical patterns in PCB distribution, aiding in the identification of contamination sources and the development of targeted remediation strategies. By isolating the dominant factors influencing PCB variability, policymakers and environmental managers can implement more effective interventions to reduce contamination and mitigate associated risks to human and ecological health. For example, targeting industrial discharge and improving wastewater management could significantly reduce PCB levels in aquatic systems (Gao et al. 2017; Zhao et al. 2019).

The component matrix in Table 8 shows the three components recovered by PCA, as well as the loadings of each variable (PCB3, PCB5, PCB17, PCB21, PCB61, PCB99, PCB98, PCB111, PCB151, PCB161, PCB159, PCB168, PCB183, PCB187, and PCB206). PCB21 (0.619), PCB99 (0.627), PCB98 (−0.962), PCB111 (−0.964), PCB168 (0.330), PCB183 (0.646), PCB187 (0.593), and PCB206 (0.991) have considerable positive loadings in Component 1. This indicates a close link between Component 1 and these variables. High positive loadings for PCB151 (0.960) and PCB161 (0.925) in Component 2 suggested a strong relationship between these variables. Component 3 is distinguished by strong positive loadings for PCB159 (0.765) and PCB168 (0.782), as well as lower loadings for certain variables. These loadings help us understand the variables that influence each component in unique ways. Component 1, for example, appears to be associated with multiple variables, including PCB21, PCB99, PCB98, and others. Component 2 is primarily controlled by PCB151 and PCB161, however, PCB159 and PCB168 have an impact on Component 3. In general, this component matrix helps to expose the data's main structure as well as the relationships between the original variables and the derived PCA components.

The three components were found using PCA with Varimax rotation and Kaiser Normalization. Table 9's rotational component matrix displays the loadings of each variable (PCB3, PCB5, PCB17, PCB21, PCB61, PCB99, PCB98, PCB111, PCB151, PCB161, PCB159, PCB168, PCB183, PCB187, and PCB 206) on the three components. Component 1 is distinguished by strong positive loadings of PCB3 (0.869), PCB21 (0.981), PCB61 (0.831), PCB187 (0.966), and other variables. This component reflects a set of chemically linked components. Component 2 contains a high positive loading of PCB99 (0.935), PCB161 (0.547), PCB151 (0.485), PCB183 (0.510), and other compounds. This component reflects the variance of these variables. PCB168 (0.876), PCB17 (−0.806), and PCB183 (0.852) all have high positive loadings in component 3. These components have a significant impact on component 3. The varimax rotation improves component readability by raising the variance of the loadings within each component while limiting component overlap. This rotation strategy facilitates in the detection of significant patterns in data. To summarize, the rotated component matrix gives a better understanding of how each variable affects the components obtained after rotation. Three components – Component 1, Component 2, and Component 3 − appear to be linked to distinct PCB chemicals, variations in other compounds, and different sets of factors. This rotated component matrix can assist in clarifying and comprehending the data's underlying structure.

The contributions of individual PCB congeners varied significantly across the study locations, reflecting the influence of local environmental conditions and pollution sources (Figure 2).

Across all three locations, higher chlorinated congeners (e.g., penta-, hexa-, and hepta-PCBs) exhibited the highest percentage contributions. For instance, hexa-PCBs accounted for 45% of the total PCB load in the Forcados River, indicating their dominance. This trend aligns with findings from previous studies, which highlight the persistence and bioaccumulation potential of higher chlorinated congeners in aquatic systems due to their hydrophobicity and resistance to degradation (Van den Berg et al. 2006; Gao et al. 2018). These congeners are more likely to adhere to sediments and biota, leading to long-term environmental persistence.

• Forcados River: The dominance of hexa-PCBs and hepta-PCBs in this location underscores the significant influence of industrial emissions and oil-related pollution. These congeners are often associated with petroleum-derived products and are likely introduced through activities such as oil spills, industrial discharges, and jetty operations. The proximity of the river to industrial hubs supports this observation (Zhao et al. 2019; Paschal et al. 2022).

• Lake Eleyele: While hexa-PCBs also contributed significantly (40%) in Lake Eleyele, there was a noticeable presence of lower chlorinated congeners, such as tri- and tetra-PCBs (25% combined). This distribution reflects the influence of agricultural runoff and urban wastewater, which often carry less chlorinated PCBs that are more water-soluble and prone to transport in surface water systems (Vorkamp 2018).

• Apete Wells: The PCB contributions in the wells showed a more even distribution, with penta-PCBs contributing 35% and tetra-PCBs contributing 25%. This broader profile likely results from mixed contamination sources, including leachates from waste disposal sites, abandoned ponds, and possible infiltration of surface pollutants. Groundwater systems may act as reservoirs for water-soluble congeners, allowing for their gradual accumulation (Gorshkov et al. 2017; Unyimadu et al. 2018).

The predominance of higher chlorinated PCBs in surface water systems (rivers and lakes) presents a significant risk of long-term contamination. These congeners are less biodegradable, tend to accumulate in sediments, and are more likely to enter the food chain, posing risks to aquatic ecosystems and human health (Van den Berg et al. 2006; Gao et al. 2018). In contrast, the presence of lower chlorinated congeners in groundwater systems (wells) highlights the potential for contamination through infiltration and transport of water-soluble compounds. This underscores the vulnerability of groundwater resources to surface activities.

These findings emphasize the need for targeted interventions based on the contamination profiles of each location:

• For the Forcados River, addressing industrial discharges and oil spills is critical to reducing hexa-PCB levels and mitigating long-term risks.

• For Lake Eleyele, improving wastewater treatment and controlling agricultural runoff can help limit the transport of less chlorinated PCBs.

• For Apete Wells, enhancing waste management practices and monitoring potential leachate sources, such as nearby waste disposal sites, is essential to prevent further contamination.

This study highlights the distribution, composition, and health risks of PCBs in surface and groundwater across three locations: Forcados River (River 3), Lake Eleyele, and Apete Wells. The findings provide critical insights into the extent of PCB contamination in these water systems and its implications for environmental health.

The Forcados River (River 3) recorded the highest PCB concentrations among the study locations (32.04 ± 0.35 mg/L), likely due to its direct connection to the Niger River, a major waterway known for its exposure to industrial and agricultural pollutants. River 3 serves as a tributary to the Niger, and the confluence point is characterized by activities such as oil exploration, jetty operations, and wastewater discharge. These activities contribute significantly to the PCB burden, as supported by previous studies documenting elevated PCB levels in parts of the Niger River basin (Eze et al. 2023). This emphasizes the importance of monitoring major waterways and implementing targeted pollution control strategies to protect downstream ecosystems and communities.

Lake Eleyele exhibited moderate PCB concentrations (10.38 ± 0.11 mg/L), reflecting contamination primarily from agricultural runoff and urban wastewater. The lake's proximity to densely populated areas and agricultural activities increases its vulnerability to pollution. The diverse sources of contamination highlight the need for integrated management practices to safeguard this vital water resource.

In contrast, Apete Wells recorded the lowest PCB concentrations (22.59 mg/L), with some PCB homologs falling BDLs. This relatively lower contamination can be attributed to the natural filtration properties of the soil, which reduce the mobility of hydrophobic PCBs. However, the presence of detectable PCB levels suggests infiltration from nearby waste disposal sites and abandoned ponds, underscoring the need for better waste management practices to protect groundwater quality.

The health risk assessment further revealed that the PCB concentrations in all water systems exceed international standards, posing significant risks to human health. Chronic exposure to PCBs, particularly through the consumption of contaminated water, may lead to endocrine disruption, immunotoxicity, and increased cancer risks. The risks are highest for communities reliant on the Forcados River and Lake Eleyele, where the dominance of higher chlorinated PCBs exacerbates the potential for bioaccumulation and biomagnification in the food chain.

This study underscores the urgent need for comprehensive surveillance and mitigation efforts to address PCB contamination in Nigeria's water systems. For the Forcados River, reducing industrial and oil-related discharges is critical. Lake Eleyele requires better agricultural runoff control, while Apete Wells necessitates improvements in waste disposal practices. These measures, coupled with public health awareness campaigns and stricter enforcement of environmental regulations, will help reduce PCB exposure and protect both ecosystems and human health.

The total level of PCBs identified in well and river samples from the states of Delta and Oyo exceeded the EPA's water quality standard of 0.0005 ppm. The findings call for a detailed examination of surface water and the migration of contaminants into these rivers in order to establish the likelihood that PCBs would enter through leaks and runoff. However, ongoing bio-monitoring of PCB's effects on local ecosystems is essential.

We thank the university management for providing the enabled environment for carrying the research.

We give our consent to publish the manuscript

All relevant data are included in the paper or its Supplementary Information.

The authors declare there is no conflict.