Contaminant of Emerging Concerns in Modder River Catchment of Free State: Implication for Environmental Risk and Water Sources Protection

Abstract

This study was aimed at monitoring the occurrence and potential sources of emerging contaminants in water sources within the Modder River catchment. Selected water quality indicators were analysed by Hanna multi-parameter meters. Emerging contaminants such as acetaminophen, carbamazepine, ibuprofen, atrazine, simazine, metolachlor, terbuthylazine, 17-alpha-ethinyl-estradiol, estradiol, progesterone, and testosterone were analysed by high performance liquid chromatography mass spectrometry. The sources of emerging contaminants were determined by statistical methods such as Pearson correlation and hierarchical cluster analysis. Results showed that all the sampled water sources have some level of questionable drinking water quality and necessitate some amount of treatment to reduce the contamination before consumption, especially DO, EC, and pH. The 17-alpha-ethinyl-estradiol mean values in rivers (7.79 and 31.55 µg/L), dams (1.83 and 6.90 µg/L), and treated drinking water (0.2 and 0.73 µg/L) were the highest in summer and autumn seasons, respectively. Wastewater effluents, domestic sewage, urban surface runoff, agricultural runoff, and illegal dumping were identified as the possible sources of emerging contaminants pollution. Waste management education, proper application of herbicides, and advance wastewater treatment methods were some of the suggested mitigation strategies. The outcomes may be relevant for environmental protection and water sustainability in the catchment.

1. Introduction

Water is a valuable and limited resource that is essential for the survival of all known life forms [1,2]. Dams, rivers, streams, and lagoons are important water sources for both human welfare and aquatic life [1]. They are critical for domestic use, agricultural endeavours, industrial growth, leisure, and aquatic habitation and reproduction [3]. The majority of urban areas were built on or near freshwater sources [2]. As a result of industrialisation and population growth, various pollutants such as emerging contaminants from manufacturing, wastewater treatment works, agricultural practises, and household tasks find their way into urban water sources, contaminating and degrading their quality [4]. Many urban areas have high levels of production and consumption of emerging contaminants as a result of economic development and population growth [5]. The presence of emerging contaminants (such as steroid hormones, personal care products, pesticides, and pharmaceuticals) in water sources has become widespread and of great concern in recent years due to their high potency and specificity of undesirable ecological effects when interacting with biological systems and human society [6]. Exposure to these contaminants may cause acute or chronic effects, such as carcinogenic, mutagenic, and teratogenic effects, depending on their concentration and exposure level. Furthermore, chemical mixtures of emerging contaminants in water sources can be extremely hazardous to human and aquatic life [7]. Ma et al. [3] stated that their presence in water sources not only endangers aquatic and human life but also impedes public and monetary development.

South Africa currently uses mostly surface water, with groundwater accounting for only 15% of total water consumption [8]. The land use in the Modder River catchment is primarily urban and irrigated cultivation. Bloemfontein is the only significant development in the Modder River catchment [9], while farming is dominated by dairy, mixed farming, and sheep farming. The Modder River catchment is important in supplying water for domestic, agricultural, and industrial use in Bloemfontein and surrounding townships. Tsokeli [10] claims that the Modder River is being used to its full potential. Oke and Alowo [11] went on to say that there have been changes, trends, and threats to water resources in the Modder River catchment. They mentioned population growth and urbanisation as examples of such changes. Changes such as population growth and urbanisation have been reported to cause pollution, causing water sources to gradually lose their original quality state, hydro-morphological characteristics, and be accumulated by concentrations of pollutants (e.g., emerging contaminants), which may have negative impact on the aquatic ecosystem [12] and threaten the sustainability of high-quality water supplies [13]. Moreover, the World Health Organization (WHO) reported that the countries with the most limited access to drinking water are primarily those in Sub-Saharan Africa, including South Africa [14]. Water resources in South Africa are scarce and unevenly distributed. It is expected that competition for water among users will intensify in the near future. As a result, monitoring water quality and emerging contaminants pollution in water sources is critical in preventing water pollution and implementing corrective measures [3].

While the number of studies investigating and monitoring the occurrence of emerging contaminants in South African water environments has increased [15,16,17,18,19,20], there is still a dearth of studies monitoring emerging contaminants in the Modder River catchment in the Free State province. Furthermore, water sources face substantial risks from both point and non-point sources pollution, which act as major pathways for the enrichment of emerging contaminants in the Modder River catchment. However, there are no published studies determining the sources of emerging contaminants in the Modder River catchment. Multivariate statistical methods have become widely used in pollution source apportionment in recent years [4,13,21,22,23,24]. They are useful in determining the source of chemical contaminants in water environments [3], which in turn, provide valuable insights into effective preventative measures and controls against emerging contaminants pollution [25]. Therefore, this study, which is the first in the Modder River catchment in the Free State province, aims to analyse the water quality indicators and quantify the level of emerging contaminants in various water sources. It also aims to determine the sources of emerging contaminants pollution using two statistical methods, namely, Pearson correlation and hierarchical cluster analysis. The outcomes of this research may be crucial in inhibiting water pollution from emerging contaminants for the protection and sustainability of the Modder River catchment in the Free State province of South Africa.

2. Materials and Methods

2.1. Study Location

The study area, which is the Modder River catchment (Figure 1), is located in the central part of South Africa. It is precisely positioned at 29°2′25″ S and 24°37′42″ E in the southwestern part of the Free State Province [26]. The summer climate is moderate to hot, with an average minimum temperature of 12 °C and a maximum temperature of 30 °C. The average minimum temperature in winter is 3 °C, with a maximum temperature of 18 °C. This area is in a summer rainfall region, with rainfall ranging from 300 mm in the west to 550 mm in the east. Early September to mid-April is the rainy season, with a dry winter. The catchment is part of the Upper Orange Water Management Area [9,11]. It has a total area of approximately 17,366 km² and is 1200 m above sea level on average. The Modder River was traditionally a seasonal river, as are most inland rivers in South Africa, but due to the construction of three significant dams, namely the Rustfontein, Mockes, and Krugersdrift Dams, the river now resembles a permanent river [11]. The catchment is characterised by very shallow slopes and easy pooling of water, which influences the attenuation period of floods and high flow conditions. The vegetation is mostly grassland [9]. Extensive animal farming is noticeable in some parts of the catchment. Some dry land cultivation occurs where precipitation and soil conditions are favourable, with significant areas irrigated below the main storage dams [27]. The Mangaung Metropolitan Municipality (MMM) controls the city of Bloemfontein, which is the only significant development in the Modder River basin [9]. A number of rivers are thought to be part of the Modder River catchment, including the Modder (Likatlong), Renosterspruit, Kleinmodder, Koringspruit, and Bloemspruit. Moreover, various dams, including the Rustfontein, Mockes, Maselspoort, Krugersdrift, and Seroalo dams, also form part of the catchment. These water sources supply water for domestic, agricultural, and industrial use across the area [28].

2.2. Collection of Water Samples

The study focused only on the most accessible rivers, dams, and treated drinking (tap) water sources. During the summer (December 2023) and autumn (April 2024) seasons, 24 samples were collected using the grab sampling method from sources such as treated drinking water, rivers, and dams (Figure 1). Every season, twelve (12) samples were taken: five (5) from dams (SWMSO1, SWMD02, SWKD08, SWSD09, and SWRUSD10), five (5) from rivers (SWRS03, SWBS04, SWKOR05, SWMOR06, and SWKLM07), and two (2) from treated drinking (tap) water (TWWRUS01 and TWWMSP02), as shown in Supplementary Table S1. A 750 mL clean glass bottle was used to collect each sample during the sampling campaign. In dams and rivers, water was collected against the flow direction. Before collecting tap water, the tap was left running for a few minutes. The labelled samples were stored in a cooler box filled with ice until they were transported to the laboratory for analysis. All samples were refrigerated at 4 °C after arrival at the laboratory until extraction and analysis.

2.3. Reagents and Materials

Chemical standards of high purity (>98%) for acetaminophen, carbamazepine, ibuprofen, triclosan, atrazine, terbuthylazine, simazine, metolachlor, 17-alpha-ethinyl-estradiol, estradiol, progesterone, and testosterone were procured from Sigma Aldrich (St. Louis, MO, USA) and Dr Ehrenstorfer (Augsburg, Germany). High-performance liquid chromatography (HPLC) grade methanol (MeOH), ammonium hydroxide, acetonitrile, and formic acid were also purchased from Sigma Aldrich (St. Louis, MO, USA). Ultra-pure water (99%) used during experiments was obtained from Millipore (Burlington, MA, USA). The solid phase extraction cartridges (Strata C18, 6 mL) were obtained from Phenomenex (Torrance, LA, USA). The stock solutions for individual standards were prepared using 1 μg/L methanol. After preparation, the stock solutions were stored in amber glass bottles at 4 °C.

2.4. Sample Preparation and Extraction

The methods of sample extraction for targeted compounds, namely, acetaminophen, carbamazepine, ibuprofen, triclosan, atrazine, terbuthylazine, simazine, metolachlor, 17-alpha-ethinyl-estradiol, estradiol, progesterone, and testosterone, were modified from methods reported by Odendaal et al. [29] and Mugudamani et al. [30]. Briefly, particulate matter was eliminated from the samples (750 mL) by filtering them via glass fibre filters. The ultra-pure water (99%), spiked with emerging contaminants chemical standards, was used to improve solid phase extraction parameters. The SPE cartridges were equilibrated with a 6 mL of methanol and Milli-Q water before extraction. Afterward, samples were loaded at a flow rate of approximately 6 mL/min. The sample cartridges were cleaned with 6 mL of purified water and dried for 20 min under vacuum. Then, 2 mL methanol was used to slowly elute the bound sample from the dried cartridges, followed by 2 mL ethyl acetate. Using a Thermo Scientific Savant Speedvac SC 210A concentrator (Waltham, MA, USA), the eluant was vacuum dried until nearly dry. Then, 0.1% formic acid and 1 mL of filtered water were used to reconstitute the extract.

2.5. Sample Analysis

The analysis was performed on a high-performance liquid chromatograph linked to a hybrid triple quadrupole ion trap mass spectrometer (ABSCIEX 4000, Framingham, MA, USA). The software utilised for all data processing and acquisition was Analyst 1.5 (AB SCIEX). To analyse the samples, both positive and negative ionisation techniques were employed. In positive ionisation mode, 20 microlitres (μL) of each extracted sample was separated on a C18 (150 mm × 4.6 mm, Gemini NX, Phenomenex, Torrence, CA, USA) column at a flow rate of 300 μL/min using a 5 min gradient from 5% solvent A (H₂O/0.1% formic acid) to 95% solvent B (MeOH/0.1% formic acid) with a total run time of 9 min to allow for column re-equilibration. Eluting analytes were electrospray ionised in the TurboV ion source with a heater temperature of 500 °C to evaporate excess solvent, 40 kPa nebuliser gas, 40 kPa heater gas, and a curtain gas of 15 kPa. The ion spray voltage was set at 5500 V. In negative ionisation mode, 20 μL of each extracted sample was separated on a C18 (150 mm × 4.6 mm, Gemini NX, Phenomenex) column at a flow rate of 300 μL/min using a 2 min gradient from 5% solvent A (H₂O/0.1% NH₃OH) to 95% solvent B (MeOH/0.1% NH₃OH) with a total run time of 10 min to allow for column re-equilibration. Eluting analytes were electrospray ionised in the TurboV ion source with a heater temperature of 500 °C to evaporate excess solvent, 40 kPa nebuliser gas, 40 kPa heater gas, and a curtain gas of 15 kPa. The ion spray voltage was set −4500 V. Multiple reaction monitoring transitions were used for each target analyte throughout analysis. Multiple reactions monitoring (MRM) provides increased selectivity and reduces the likelihood of spectral interferences. The peak area on the chromatogram produced by the first and most sensitive transition was known as the qualifier, while the peak area produced by the second transition was known as the quantifier. The retention period for these two transitions had to match, and the qualifier acted as an extra layer of verification for the analytes’ existence [29,30]. The details of the MRM transitions for each compound are listed in Table S2. The first transition values (m/z) were used in this study as bolded in Supplementary Table S2.

2.6. Quality Control

For quality control and assurance, selectivity, linearity, and limit of quantification were considered in this study. Samples were submitted in batches with solvent blank runs between each sample analysed and quality control samples of known concentration interspersed. For each analyte, a four-point calibration curve with a linear fit through the origin was generated, ranging in concentration from 0.001 ppm to 1 ppm. The selectivity was evaluated by comparing the retention time of the extracted blank sample and the retention time of each spiked standard or IS in the extracted samples. If any interferences were observed, the signal of their peak area should be less than 20% of the peak area of LOQ and less than 5% of the average peak area of IS in blank samples. The calibration values showed a concentration range of 5 × 10⁻⁵ to 1 × 10⁻¹ mg/L. There was no interference effect from the endogenous matrix at the retention time for all analytes and IS. The linear fit yielded a correlation coefficient (r) value exceeding 0.98. Furthermore, the quantification limits ranged from 0.0001 mg/L to 0.1 mg/L as presented in Supplementary Table S3 [30].

2.7. Water Quality Indicators in the Modder River Catchment

The physicochemical parameters of various water sources targeted in the Modder River catchment were measured on the spot with portable multi-parameter probes. Temperature, electrical conductivity, total dissolved solids, and pH were measured using a Hanna HI 98195 (Hanna Instruments, Inc., Smithfield, RI, USA) while dissolved oxygen was determined using a Hanna HI98193 (Hanna Instruments, Inc., Smithfield, RI, USA) multi-parameter instrument as presented in Supplementary Table S4.

2.8. Statistical Analysis and Sources Apportionment

The data were analysed using Microsoft Excel. The quantified water quality indicators and emerging contaminants concentrations in rivers, dams, and treated drinking water samples were used to compute descriptive statistics such as minimum (Min), maximum (Max), Mean, and Standard deviation (SD). Matlab software R2023a (The MathWorks, Inc, MA, USA) was used to perform statistical analyses such as Pearson correlation and hierarchical cluster analysis. These multivariate statistical techniques were used to identify the sources of emerging contaminants pollution in water sources within the Modder River catchment.

3. Results and Discussion

3.1. Analysis of Water Quality Indicators in the Modder River Catchment

Water quality indicators such as dissolved oxygen (DO), electrical conductivity (EC), pH, total dissolved solids (TDS), and temperature, presented in Supplementary Table S5, were considered as they can influence the taste and feel of water and affect the rate of corrosion and scale build-up of water channels [2,31]. In rivers, an average values of 7.98 and 8.10 were recorded for pH in summer and autumn seasons, respectively (Supplementary Table S5). pH average values of 8.03 and 8.08 in summer and autumn, respectively, were witnessed in dams. Treated drinking water showed average pH values of 4.33 and 8.29 in summer and autumn seasons, respectively. The pH average values in all water sources were within the tolerable values of SANS 241 [32] and WHO [33] guidelines in all seasons (Supplementary Table S5). However, during the summer season in treated drinking water, the pH value was lower than the acceptable values of SANS 241 [32] and WHO [33] guidelines. The low average value of pH in treated drinking water during the summer season suggested that the nature of the water was acidic and may lead to slow erosion of tooth enamel among water consumers [34]. Moreover, the low pH detected during summer may also suggest that treated water was polluted. Considering that water samples were collected from tap water from different treatment plants, household plumbing (tap) corrosion and water treatment plant decomposition may have influenced the low level of pH during summer. The values of pH in river and dam samples during all seasons confirms that the nature of water in this study was alkaline, suggesting that the surrounding community, aquatic organisms, and animals are safe from adverse health effects. According to Gintamo et al. [35], if pH values are higher than the acceptable limit, human health may be endangered, mostly from effects that occur in the eyes, nose, mouth, and abdomen. Shoeb et al. [36] also supported this finding by stating that pH out of the acceptable limits may deleteriously threaten aquatic diversity and reproduction of organisms.

The temperature average values in rivers were 35.16 and 21.21 °C in summer and autumn seasons, respectively (Supplementary Table S5). In dams, they were recorded as 35.26 and 22.92 °C while in treated drinking water they were 32.17 and 26.36 °C in summer and autumn, respectively. The results showed variation in temperature average values across all water sources; this might be attributed to the sampling time, which ranged from morning to late afternoon. In support of this statement, Khan and Wen [37] mentioned that the increase in solar radiation from the morning to midday may be the reason for variation in water temperatures. The mean values of temperature in all water sources during summer season were higher than the 25 °C recommended by SANS 214 [32], as presented in Supplementary Table S5. These outcomes correspond to the findings of Birajdar et al. [31] in India. Temperature is important as it influences the chemistry in water. Water with higher temperatures can dissolve more minerals from the surrounding rock and will therefore have a higher electrical conductivity. Furthermore, warm water holds less dissolved oxygen than cool water, and may not contain enough dissolved oxygen for the survival of different aquatic species [38,39].

The total dissolved solids (TDS) average concentrations in rivers were 175.6 and 229.2 mg/L in summer and autumn seasons, respectively (Supplementary Table S5). In dams they were 137.4 and 114.0 mg/L in summer and autumn, respectively. The TDS average values of 52.2 and 103.0 mg/L were recorded in treated drinking water during summer and autumn seasons, respectively. All the water sources showed average values of TDS within the acceptable standard set by SANS 214 [32] and WHO [33] guidelines, as presented in Supplementary Table S5. Similar findings were also reported by other authors [31,40,41]. The total dissolved solid in water sources indicates the salinity behaviour of water, and water sources with high TDS values indicate that water is unsuitable for drinking purpose [31]. The TDS is comprised of inorganic and organic particles in water. In water, inorganic materials may include clay or silt, while organic particles may include algae and plant fibres, among others. Soil erosion from the surrounding agricultural activities, runoff from nearby forestry activities, and urban stormwater may have influenced in the level of TDS [42]. Considering that all the water sources showed average values of TDS within the acceptable standard, the water may be usable and less harmful to the local community [31]. This also shows that there was no significant pollution from soil erosion from the surrounding agricultural activities, runoff from nearby forestry activities, and from urban storm water. Water with total dissolved solids above the acceptable values of <1200 mg/L [32] and 500 mg/L [33] may cause irritation, fluorosis, and gastrointestinal toxicities among consumers [35]. Moreover, water containing high amounts of undissolved solids may cause laxative or constipation effects [41].

The average values of dissolved oxygen (DO) in rivers were recorded as 2.55 and 2.94 mg/L in summer and autumn seasons, respectively (Supplementary Table S5). In dams, the DO mean values were 2.92 and 3.54 mg/L while in treated drinking water they were 2.37 and 2.49 mg/L in summer and autumn seasons, respectively. The average values of dissolved oxygen in all water sources in all seasons were less than 5 mg/L, showing that the water bodies in the Modder River catchment did not having enough oxygen. DO is the amount of oxygen in aquatic environments that is accessible to fish, invertebrates, and all organism in the water. Low dissolved oxygen may be harmful to aquatic organisms [37], and species such as fish cannot survive for long in water with dissolved oxygen less than 5 mg/L [36,43,44]. When oxygen is low, nutrients bound to the bottom sediments can be released into the water column, thereby permitting more plankton growth and eventually more oxygen depletion [43]. Moreover, the low level of dissolved oxygen in water is a sign of contamination, probably from nearby industrial sites, agricultural farms, and wastewater discharge, which may release high organic matter content that may lead to lack of oxygen in the water.

Electrical conductivity (EC) showed average values of 351.40 and 479.20 μS/cm in rivers during summer and autumn seasons, respectively (Supplementary Table S5). The EC mean values in dams during summer and autumn were as 274.20 and 66.41 μS/cm, respectively. Moreover, treated drinking water had EC mean values of 95.67 and 207.50 μS/cm in summer and autumn, respectively. The average values of electrical conductivity in rivers and dams during all seasons were higher than the acceptable limits set by SANS 214 [32] and within the WHO [33] acceptable limit (Supplementary Table S5). In treated drinking water, they were higher than the acceptable standards set by SANS 214 [32] during autumn and within the WHO [33] acceptable limits in all seasons. Makhadi et al. [45] indicted that high levels of electrical conductivity are a result of high concentrations of ionic constituents present in the water. Khan and Wen [37] also detailed that streams that run through areas with clay soils tend to have higher conductivity because of the presence of materials that ionise when washed into the water; this is true in this study, as the geology of the area is characterised by siltstones interbedded with layers of clay, sandstone, and shale [45]. As a result of high electrical conductivity, saline conditions may be observed [36], thus affecting the taste and freshness of water.

3.2. Emerging Contaminants in the Modder River Catchment

A total of 24 samples were collected during summer (n = 12) and autumn (n = 12) seasons from rivers (n = 5), dams/reservoirs (n = 5), and treated (tap) drinking water (n = 2). During laboratory analysis, twelve organic contaminants were targeted. These contaminants included pharmaceuticals (acetaminophen, ibuprofen, and carbamazepine), personal care products (triclosan), herbicides (atrazine, simazine, terbuthylazine, and metolachlor), and steroid hormones (17-alpha-ethinyl-estradiol, estradiol, progesterone, and testosterone). They were chosen primarily for their frequency of detection in water sources, availability of reagents, and reported potential risks on living organisms. The occurrence of selected contaminants in each sampling site are presented in Supplementary Table S6, while descriptive statistics of the twelve (12) selected contaminants in various water sources are presented in

Table 1. Emerging contaminants in various water sources within the Modder River catchment.

Compound	Summer (n = 12)			Autumn (n = 12)
	DF	Min-Max	Mean/±SD	DF	Min-Max	Mean/±SD
Rivers (n = 5)
Acetaminophen (µg/L)	0	<LOQ	<LOQ	0	<LOQ	<LOQ
Carbamazepine (µg/L)	4	0.40–1.43	0.7 ± 0.49	4	<LOQ–0.32	0.25 ± 0.07
Ibuprofen (µg/L)	4	0.38–2.11	1.16 ± 0.76	3	<LOQ–2.26	0.92 ± 1.16
Triclosan (µg/L)	0	<LOQ	<LOQ	0	<LOQ	<LOQ
Atrazine (mg/L)	5	0.03–0.11	0.05 ± 0.03	5	0.0014–0.03	0.69 ± 0.01
Metolachlor (mg/L)	5	0.02–0.12	0.06 ± 0.04	5	0.0007–0.03	0.02 ± 0.01
Simazine (mg/L)	4	0.10–3.22	1.041 ± 0.48	2	<LOQ–0.91	0.69 ± 0.30
Terbuthylazine (mg/L)	5	0.03–0.14	0.08 ± 0.05	5	0.0016–0.03	0.01 ± 0.01
17-alpha-ethinyl estradiol (µg/L)	5	1.08–14.5	7.79 ± 5.34	4	<LOQ–53.80	31.55 ± 18.79
Estradiol (µg/L)	0	<LOQ	<LOQ	0	<LOQ	<LOQ
Progesterone (µg/L)	0	<LOQ	<LOQ	0	<LOQ	<LOQ
Testosterone (µg/L)	0	<LOQ	<LOQ	0	<LOQ	<LOQ
Dams/reservoirs (n = 5)
Acetaminophen (µg/L)	0	<LOQ	<LOQ	1	<LOQ–0.09	-
Carbamazepine (µg/L)	5	0.01–0.21	0.12 ± 0.08	5	0.03–0.19	0.08 ± 0.06
Ibuprofen (µg/L)	0	<LOQ	<LOQ	2	<LOQ–0.03	0.02 ± 0.01
Triclosan (µg/L)	0	<LOQ	<LOQ	0	<LOQ	<LOQ
Atrazine (mg/L)	5	0.02–0.04	0.03 ± 0.01	5	0.003–0.02	0.01 ± 0.01
Metolachlor (mg/L)	5	0.01–0.04	0.03 ± 0.01	5	0.01–0.02	0.01 ± 0.01
Simazine (mg/L)	4	<LOQ–0.24	0.13 ± 0.11	4	<LOQ–0.29	0.12 ± 0.12
Terbuthylazine (mg/L)	5	0.02–0.08	0.05 ± 0.02	5	0.003–0.03	0.01 ± 0.08
17-alpha-ethinyl estradiol (µg/L)	5	0.25–3.40	1.83 ± 1.13	5	1.30–14.80	6.90 ± 5.08
Estradiol (µg/L)	0	<LOQ	<LOQ	0	<LOQ	<LOQ
Progesterone (µg/L)	0	<LOQ	<LOQ	0	<LOQ	<LOQ
Testosterone (µg/L)	0	<LOQ	<LOQ	0	<LOQ	<LOQ
Treated (tap) drinking water (n = 2)
Acetaminophen (µg/L)	0	<LOQ	<LOQ	0	<LOQ	<LOQ
Carbamazepine (µg/L)	0	<LOQ	<LOQ	0	<LOQ	<LOQ
Ibuprofen (µg/L)	0	<LOQ	<LOQ	0	<LOQ	<LOQ
Triclosan (µg/L)	0	<LOQ	<LOQ	0	<LOQ	<LOQ
Atrazine (mg/L)	2	0.02–0.12	0.07 ± 0.07	2	0.01–0.03	0.02 ± 0.01
Metolachlor (mg/L)	2	0.01–0.09	0.05 ± 0.06	2	0.004–0.04	0.02 ± 0.01
Simazine (mg/L)	1	<LOQ–0.04	-	0	<LOQ	<LOQ
Terbuthylazine (mg/L)	2	0.02–0.16	0.09 ± 0.10	2	0.004–0.02	0.01 ± 0.003
17-alpha-ethinyl estradiol (µg/L)	2	0.14–0.26	0.2 ± 0.08	2	0.42–1.04	0.73 ± 0.40
Estradiol (µg/L)	0	<LOQ	<LOQ	0	<LOQ	<LOQ
Progesterone (µg/L)	0	<LOQ	<LOQ	0	<LOQ	<LOQ
Testosterone (µg/L)	0	<LOQ	<LOQ	0	<LOQ	<LOQ

Notation: n = number of samples; LOQ = limit of quantification; mg/L = milligram per litre; DF = detection frequency; min = minimum; max = maximum; ±SD = standard deviation.

. They were summarised as detection frequency (DF), min, max, mean, and standard deviations (±SD). The mean concentrations of contaminants detected only in one sample were not calculated. Generally, their concentrations ranged from <LOQ to 7.79 mg/L in the summer season to <LOQ to 31.55 mg/L in the autumn season. The occurrence of emerging contaminants within the Modder River catchment are presented and discussed based on their source of occurrence.

3.2.1. The Presence of Emerging Contaminants in Rivers

Among the pharmaceutical and personal care products (n = 5), the respective detection rates of acetaminophen, carbamazepine, ibuprofen, and triclosan during summer were 0, 4, 4, and 0, while in autumn they were 0, 4, 3, and 0. The detection rates of carbamazepine in this study are similar to those reported in rivers in Cotonou, Benin; Mbarara, Uganda; and Harare, Zimbabwe [20]. Their respective mean concentrations in summer and autumn seasons are presented in Table 1. The highest concentrations among the pharmaceuticals and personal care products detected in rivers were from ibuprofen, which had high average concentrations in both the summer and autumn seasons. The occurrence of ibuprofen in this study coincides with those reported in the Mississippi River, USA [46], Hanoi, Vietnam [47], Chongqing, China [48], the Msunduzi River, South Africa [49], and the Apies River, South Africa [50]. Moreover, similar cases of the occurrence of carbamazepine in rivers were also reported by other researchers [16,20,51,52,53,54]. The majority of streams in this study run via townships, the city centre, and agricultural farms. Illegal waste dumping near rivers was visible in the majority of townships. Ibuprofen is used to treat pain and fever, which are common illnesses in many communities [55]. Given that ibuprofen is used to treat a variety of diseases and is available without a prescription, it was not surprising that it was the leading pollutant in rivers. The presence of carbamazepine as a prescription drug may have been influenced by the presence of patients suffering from seizures and bipolar disorder in those areas [56]. Waste dumped near rivers that run through these settlements may contain unwanted or expired medical drugs such as ibuprofen and carbamazepine, introducing these contaminants into the rivers. Some of the emerging contaminants may be spread into water bodies by wastewater treatment works that discharge their effluents into nearby streams. Furthermore, farmhouses equipped with septic tanks and latrine toilets may have influenced the concentration of these pharmaceutical compounds in nearby streams.

The number of samples with positive detection of the pesticide’s atrazine, metolachlor, simazine, and terbuthylazine in river samples (n = 5) were 5, 5, 4, and 5 in the summer season and 5, 5, 2, and 5 in the autumn season. Their corresponding average values during the summer season and autumn season are presented in Table 1. Simazine had the highest mean concentrations in all seasons, which is in line with the study conducted in rivers around Mangaung Metropolitan Municipality, South Africa [30]. Other researchers also reported the occurrence of atrazine in Kabwe, Zambia [57], metolachlor in the Samambaia River basin, Brazil [58], and terbuthylazine in Western Cape, South Africa [59]. The presence of herbicides in rivers was not surprising in this catchment, particularly trazine herbicides such atrazine, simazine, and terbuthylazine, which were the most prevalent herbicides in river samples. Trazine herbicides are thought to be effective and inexpensive chemicals that are mostly employed in crop management [60]. The agriculture sector is the foundation of the Free State province’s economy. It is well-known for its abundant production of crops (maize, sunflower, nuts, etc.). All these tasks require the use of weed control herbicides before, during, and after cultivation. Moreover, Bloemfontein is the only significant development in the Modder River catchment [9]. Bloemfontein’s well-maintained highways, public parks, industrial zones, and golf clubs are its defining features. Herbicides are used extensively in these key locations to manage weeds in pavers, parks, golf courses, roadside vegetation, buildings, and industrial areas. Herbicide concentrations in adjacent streams may rise due to runoff from neighbouring highways, public spaces, golf courses, industrial zones, and agricultural fields.

Among the steroid hormones, the rates of detection for 17-alpha-ethinyl-estradiol, estradiol, progesterone, and testosterone were 5, 0, 0, and 0 during the summer season and 4, 0, 0, and 0 during autumn. The mean concentrations of these contaminants are presented in Table 1. The concentrations of 17-alpha-ethinyl-estradiol were the highest in all seasons. However, the autumn season showed the highest concentration overall. Occurrences of 17-alpha-ethinyl-estradiol in rivers has been reported around the world: in Ningbo, China [12], the Atibaia River, Brazil [61], Lagos, Nigeria [62], Southern Brazil [63], and the Liao River Basin, Northen China [64]. The presence of steroid hormones in rivers within this catchment may have been influenced by wastewater effluents that are discharged into nearby streams. Wastewater treatment works in this area receive influent from domestic, industrial, and hospital sources, thus acting as major sources of these compounds [65]. In addition, one of the wastewater treatment works within the study area was out of order. Hence, wastewater influents from industrial activities e.g., pharmaceutical manufacturing, and domestic and hospital sources were received and discharged into nearby streams without any treatment. This situation might have influenced the high concentration of this compound in river samples, which is the highest ever detected in rivers. Rivers that pass through townships and city centres are likely to be contaminated by these compounds as a result of illegal dumping of household wastes (e.g., medical drugs) near them. According to Forghani et al. [65], estradiol is a natural estrogenic hormone released by humans and livestock. Animals that drink the water from these rivers, as well as people who visit them, may end up introducing these hormones into the river water through open urination and defecation. Moreover, runoff from animal husbandry activities near some rivers may also introduce these hormones. In summation, the emerging contaminants concentrations were highest in the autumn season, revealing that the autumn season is the largest contributor to river pollution in the Modder River catchment. The contamination was mostly caused by 17-alpha-ethinyl-estradiol, simazine, and ibuprofen, which make them contaminants of concern in this catchment and should be put on the watch list in future pollution monitoring studies.

3.2.2. The Presence of Emerging Contaminants in Dams

The frequency of detection for the twelve (12) emerging contaminants varied significantly in dams (n = 5) within the Modder River catchment. Among them, the most notable detection frequencies were for carbamazepine, atrazine, metolachlor, terbuthylazine, and 17-alpha-ethinyl-estradiol, which were detected in all collected samples during both the summer and autumn seasons. Simazine was detected in four (4) of the five (5) collected samples in all seasons. In the Hartbeespoort Dam, South Africa, triazine herbicides such as atrazine and terbuthylazine were also detected in more than 80% of the samples, which corresponds to the outcomes of this study [66]. Other contaminants were detected in no more than two samples (Table 1). Their mean concentrations are presented in Table 1. Ibuprofen recorded the highest mean concentration in both the summer and autumn season. Various researchers have reported the occurrence of ibuprofen in surface water around the world [63,67,68]. The high concentrations of ibuprofen in dams can be explained by the fact that analgesics and non-steroid anti-inflammatory drugs are available without a prescription and are used to treat pain and inflammation, which are symptoms of a variety of diseases [56]. Ibuprofen is a medication that is used to manage and treat inflammatory diseases, rheumatoid arthritis, pain, fever, and dysmenorrhea [55]. Anticonvulsants, such as carbamazepine, are regarded as a prescription drugs that work in the brain and nervous system to control seizures, pain, and bipolar disorder [56,69]. The presence of carbamazepine in more than 80% of dam samples indicates that carbamazepine is a commonly used drug in and around the Modder River catchment. Some of the dams in this study are close to the settlements, open to the public, and serve as resorts, conference centres, fishing spots, and aquatic sports areas. People who visit those resorts or live near these dams may have urinated near the dams or improperly disposed of these drugs. Illegal dumping of household waste containing unused or unfinished medication, as well as sewage runoff from nearby communities observed during a site visit, may have contributed to the concentrations of these compounds. Furthermore, lodging houses and some farmhouses with septic and latrine toilets near some of the dams may leak these contaminants.

As presented in Table 1, the concentrations of simazine in all seasons were the highest among the pesticides found in dams, which is in line with the findings of Mugudamani et al. [30], who reported simazine as the leading herbicide in dams of Mangaung Metropolitan Municipality, South Africa. Other researchers detected atrazine, metolachlor, and terbuthylazine in surface water [66,70,71]. High concentrations of pesticides in dams may be attributed to the fact that they are located nearby agricultural fields, which necessitate the application of these herbicides to kill weeds. For weed management and upkeep, significant amounts of pesticides are used in forestry, industrial areas, parks, golf courses, and sports fields [30]. Atrazine, metolachlor, terbuthylazine, and simazine may be introduced into dams by runoff from these locations. Herbicides can be used to suppress invasive plants in water, according to Pandey et al. [72]. Therefore, the usage of these herbicides to manage aquatic weeds like algae and submerged weeds may be the reason for their presence in dams.

Among the steroid hormones, 17-alpha-ethinyl-estradiol had the highest average concentrations in all seasons (Table 1). The autumn season recorded the highest mean concentration of this compound. The occurrences of 17-alpha-ethinyl-estradiol in this study of the Modder River catchment were also reported in Joao Goulart catchments, Brazil [63], and in Nigerian river catchments [73]. 17-alpha-ethinyl-estradiol, a synthetic oestrogenic compound, is commonly used as the active ingredient in many oral contraceptives and postmenopausal hormone therapies [69]. Many factors contributed to the detection of steroid hormones in dams, including illegal dumping of household waste containing unused or expired drugs near the dams and sewage runoff from nearby settlements, which was observed near one of the sampling points. Agricultural areas near some dams, particularly those with livestock production, may also serve as a secondary source of these compounds. Some steroid hormones may be leaked from septic tanks in some farmhouses and dams with resorts. Furthermore, rivers that receive wastewater effluents and recharge these dams may introduce steroid hormones. In summation, high levels of the emerging contaminants in dams were recorded in autumn season. It was mostly caused by 17-alpha-ethinyl-estradiol, simazine, and ibuprofen, which make then contaminants of concern in this catchment. In future surface water pollution monitoring studies, these contaminants should be prioritised in this region and in other parts of the country.

3.2.3. The Presence of Emerging Contaminants in Treated Drinking Water

Among the contaminants in treated drinking water, herbicides were mostly detected. Only atrazine, metolachlor, terbuthylazine, and 17-alpha-ethinyl-estradiol were detected in all collected samples in all seasons (Table 1). These outcomes are comparable to the findings of other researchers, who also reported atrazine and terbuthylazine as the most detected contaminants in treated drinking water [29,74]. The average values of acetaminophen, carbamazepine, ibuprofen, triclosan, atrazine, metolachlor, simazine, terbuthylazien, 17-alpha-ethinyl-estradiol, estradiol, progesterone, and testosterone are presented in Table 1. The autumn season had the highest concentrations of contaminants. Among the targeted emerging contaminants in treated drinking water, 17-alpha-ethinyl-estradiol, terbuthylazine, and atrazine average concentrations were the highest. In a study conducted in Mangaung Metropolitan Municipality, treated drinking water was also found to have concentrations of herbicides such as atrazine and terbuthylazine. Van Zijl et al. [75] reported the occurrence of 17-alpha-ethinyl-estradiol in treated drinking water around Pretoria, South Africa. The use of herbicides to manage weeds in dams used as a water source in water treatment plants may be the cause of the excessive quantities of herbicides in treated drinking water. Given that they primarily absorb wastewater effluents, rivers that recharge dams utilized as water supplies for these water treatment plants may contain traces of 17-alpha-ethinyl estradiol and pesticides. The fact that atrazine is highly detectable in water sources may potentially be related to its persistence. Since herbicides like atrazine are thought to be persistent, they are the most often found pesticide in water sources [76]. The presence of these substances in the treated drinking water further suggests that the current techniques employed in the two water treatment plants, such as sedimentation, filtration, abstraction, macro/micro sieving, coagulation, and flocculation, are unable to remove them.

Supplementary Table S7 shows atrazine mean values to be above the South African acceptable limit [77,78] in all seasons. The mean values of pesticides in all seasons were above their corresponding WHO guidelines [79], as presented in Supplementary Table S7. These outcomes are in line with the findings of Mugudamani et al. [30], who reported herbicides in treated drinking water to be above the acceptable limits. These outcomes indicate a serious health concern for water consumers within the Modder River catchment. Ingestion of pesticides can have a variety of negative impacts, including poorer immunity, hormone imbalance, reproductive system disruption, carcinogenic effects, and decreased IQ, especially in young children who are still developing physically [80]. In summation, the highest sum concentration of emerging contaminants in treated drinking water was observed in autumn season. It was mostly contributed by 17-alpha-ethinyl-estradiol, terbuthylazine, and atrazine. These compounds should be prioritised before water treatment to reduce their associated health risks.

3.3. Emerging Contaminants Pollution Sources in the Modder River Catchment

Emerging contaminants are introduced into the water environment through various sources and pathways. They can be discharged into the water environment either from a point or non-point sources [81,82]. The sources of pollution were profiled during a field survey (Supplementary Table S1). From the field survey, the main possible source of pollution for emerging contaminants in the Modder River catchment may include point source such as wastewater effluents (containing domestic, industrial, and hospital wastewater), sceptic tanks, and industrial activities. Runoff from agricultural activities (crop production and animal husbandry), household waste, and domestic sewage runoff may be considered as non-point sources. Moreover, some rivers are near the highway and pass-through settlements and urban area. Therefore, run-off from roads, settlements, and parks may be attributed as non-point sources in this study [25,81]. It should be noted that a definitive mark of a source may not exist because pollutant emissions rely on many factors that may vary significantly during the transportation process. In addition, source profiles may also be affected by physical, chemical, and biological degradation of water quality variables before they enter the receptor sites. So, the bias is reasonable between the emission profiles and loading factors [82]. To make a reasonable decision, multivariate statistical methods such as Pearson correlation and hierarchical cluster analysis were used for pollution source identification.

3.3.1. Pollution Source Identification with Pearson Correlation Analysis

To establish emerging contaminants relationships and make inference on their sources of origin, Pearson correlation analysis was performed. When the degree of correlation (r) was >0.7, it was regarded as strong, 0.5 < r < 0.7 connoted moderate correlation, and <0.5 suggested week correlations [13,83,84]. According to Zhao et al. [85] and Mugudamani et al. [84], a strong correlation between contaminants suggests a possible common sources or similar chemical behaviour.

Identification of Pollution Source in Rivers

Supplementary Table S8 shows the Pearson’s correlation matrix of the emerging contaminants in rivers during summer and autumn seasons. Generally, in all seasons there was a strong relationship between atrazine/simazine; terbuthylazine/simazine; terbuthylazine/atrazine; metolachlor/simazine; metolachlor/atrazine; metolachlor/terbuthylazine; ibuprofen/carbamazepine; 17-alpha-ethinyl-estradiol/simazine; 17-alpha-ethinyl-estradiol/terbuthylazine; 17-alpha-ethinyl-estradiol/metolachlor; 17-alpha-ethinyl-estradiol/ibuprofen; carbamazepine/simazine; carbamazepine/atrazine; carbamazepine/terbuthylazine; metolachlor/carbamazepine; and 17-alpha-ethinyl-estradiol/carbamazepine, which is an indication of similar sources of origin. The strong correlation between pesticides such as atrazine/simazine; terbuthylazine/atrazine; terbuthylazine/simazine; metolachlor/simazine; metolachlor/atrazine; and metolachlor/terbuthylazine may be attributed to agricultural activities, wastewater effluents, and runoff from roads, settlements, and urban areas. Most of the rivers are surrounded by agricultural activities producing maize, soybeans, wheat, sorghum, sunflower, etc., which demand the application of weed control herbicides for their management. Additionally, Bloemfontein is a developed location featuring industrial zones, golf clubs, public parks, and well-maintained highways [30]. Significant quantities of herbicides are used to manage weeds in paving, parks, golf courses, roadsides, buildings, and industrial sites. Therefore, runoff from agricultural fields, roads, public squares, golf courses, and industrial areas may be attributed as a possible source of these herbicides in nearby streams. There was also a strong correlation between medical drugs such as ibuprofen/carbamazepine; 17-alpha-ethinyl-estradiol/ibuprofen; and 17-alpha-ethinyl-estradiol/carbamazepine, which may be linked to leakage from farmhouse sceptic tanks, discharge of wastewater effluents into streams, runoff from animal husbandry, and dumping of household waste nearby the streams. Moreover, the strong correlation between contaminants such as carbamazepine/atrazine; carbamazepine/simazine; carbamazepine/terbuthylazine; 17-alpha-ethinyl-estradiol/atrazine; 17-alpha-ethinyl-estradiol/simazine; 17-alpha-ethinyl-estradiol/terbuthilazine; and 17-alpha-ethinyl-estradiol/metolachlor may be attributed to the fact that some of the rivers run through townships and cities, hence, animals and community members’ direct urination, runoff from clogged sewage systems, or improper waste disposal may influence the release of these contaminants in rivers. They may also be linked to discharge of wastewater. Compounds such as estradiol are natural estrogenic hormones released by the humans and livestock. Hence, direct urination from animals (e.g., goats, sheep, and cows) seen drinking river water and human beings may introduce these hormones. In this study area, most of the rivers run via city centres and residential areas (townships) with paved areas that necessitate the use of herbicides to control weeds and that are connected to sewage lines that carry all wastewater, including storm water. During one of the sampling campaigns, a clogged sewer was seen discharging into a river, possibly introducing pharmaceuticals compounds from household wastewater and herbicides carried by urban storm water. Moreover, in one of the settlement areas, a pile of domestic waste (possibly containing medical drugs) was seen adjacent to the river, which may introduce these contaminants. Additionally, because selected wastewater treatment works receive wastewater from households, industrial sources, and hospitals, their wastewater effluents discharged into these rivers may introduce traces of these contaminants.

Identification of Pollution Source in Dams

As bolded in Supplementary Table S9, the general outcomes in dams during both summer and autumn seasons revealed a strong correlation between herbicides such as atrazine-simazine; terbuthylazine-simazine; terbuthylazine-atrazine; metolachlor-atrazine; metolachlor-simazine; and metolachlor-terbuthylazine, suggesting similar anthropogenic sources. As noted by Pandey et al. [72], herbicides are used for the control of invasive plants present in water sources. Therefore, the presence of these herbicides in dams may be attributed to their use to control aquatic weeds such as algae and submerged weeds. Dams with well-kept large open spaces near them are often lawns and grasses turfed for holding picnics. Hence, herbicides may be required to suppress and control annual and perennial broadleaf and grassy weeds in those large open spaces, which may enter dams as a result of the runoff. According to Wang et al. [60], herbicides may also be applied before and after planting to control broadleaf and grassy weeds in agricultural fields. Given that the majority of the dams are surrounded by agricultural fields, pollution may be attributed to runoff from those sites. Another strong positive correlation in dams was between pairs of pharmaceutical compounds such as 17-alpha-ethinyl-estradiol/carbamazepine and 17-alpha-ethinyl-estradiol/ibuprofen, which may be linked to sewage runoff from nearby communities and leakage from septic and latrine toilets from lodging houses and farmhouses around those dams. Illegal dumping of medical waste from nearby settlements and during aquatic sports activities, conferences, or picnics may also introduce these compounds in dams. The strong positive correlation between carbamazepine/terbuthylazine and ibuprofen/atrazine may be linked to sources such as agricultural runoff and wastewater effluent released into streams that eventually discharge their water into these dams.

Identification of Pollution Source in Treated Drinking Water

As presented in Supplementary Table S10, summer and autumn seasons generally revealed the strong positive correlations of atrazine/simazine; terbuthylazine/simazine; terbuthylazine/atrazine; metolachlor/simazine; metolachlor/atrazine; metolachlor/terbuthylazine; 17-alpha-ethinyl-estradiol/simazine; 17-alpha-ethinyl-estradiol/atrazine; 17-alpha-ethinyl-estradiol/terbuthylazine; and 17-alpha-ethinyl-estradiol/metolachlor. The strong correlations between pairs of herbicides such as atrazine/simazine; terbuthylazine/simazine; terbuthylazine/atrazine; metolachlor/simazine; metolachlor/atrazine; and metolachlor/terbuthylazine may be linked to the application of herbicides to control aquatic weeds in dams used as a source of water for water treatment works. Moreover, most of these dams used as sources of water for water treatment plants are recharged by rivers that receive wastewater effluents and are surrounded by various agricultural activities (crop production and animal husbandry), which may be linked to strong correlation between 17-alpha-ethinyl-estradiol/simazine; 17-alpha-ethinyl-estradiol/atrazine; 17-alpha-ethinyl-estradiol/terbuthylazine; and 17-alpha-ethinyl-estradiol/metolachlor. Therefore, wastewater effluents, agricultural runoff, and management of aquatic weeds in dams by herbicides may eventually lead to the detection of traces of these compounds in treated drinking water [30].

3.3.2. Pollution Source Identification with Hierarchical Cluster Analysis

The hierarchical cluster analysis technique was used to determine the sources of emerging contaminants pollution in rivers, dams, and treated drinking water. According to this technique, the most likely observations fell within the same class or category [86]. In this study, emerging contaminants within the same clusters were considered as emanating from homogenous pollution sources.

Identification of Pollution Sources in Rivers

The hierarchical cluster analysis in rivers clustered the twelve targeted emerging contaminants into four clusters in all seasons, as highlighted in Figure 2 and presented in Supplementary Table S11. In summer, cluster 1 (C1) was characterised by one contaminant (simazine), which may be linked to wastewater effluent, agricultural runoff, and urban surface runoff. Cluster 2 (C2) was characterised by eight contaminants (acetaminophen, progesterone, testosterone, estradiol, metolachlor, atrazine, terbuthylazine, and triclosan), which may be linked to wastewater treatment effluent, agricultural runoff, illegal dumping, and urban surface runoff. Cluster 3 (C3) had two contaminants (ibuprofen and carbamazepine), which may be linked to wastewater treatment effluent and illegal dumping. Cluster 4 (C4) had one contaminant (17-alpha-ethinyl-estradiol), which may be linked to wastewater treatment effluent and illegal dumping. Wastewater treatment effluents discharged into rivers may contain traces of emerging contaminants as they are not designed to eliminate emerging contaminants. Illegal dumping of unused or expired medical drugs near streams may cause emerging contaminants pollution.

During the autumn season, C1 was dominated by six contaminants (acetaminophen, metolachlor, progesterone, testosterone, atrazine, and terbuthylazine), which may have been introduced by wastewater effluent, agricultural runoff, illegal dumping, and urban surface runoff. C2 was dominated by three contaminants (estradiol, carbamazepine, and simazine), which may have been introduced by wastewater effluent, agricultural runoff, illegal dumping, and urban surface runoff. Moreover, C3 had two contaminants (triclosan and ibuprofen), which are possibly a result of wastewater effluent, and C4 had one contaminant (17-alpha-ethinyl-estradiol), which probably derived from wastewater effluent and illegal dumping (Figure 2). Agricultural activities (crop production) that surround most of the rivers within the Modder River catchment may influence the occurrence of pesticides in water sources as a result of agricultural runoff. The use of herbicides to control weeds in paved areas, public parks, golf courts, and industrial areas in the city of Bloemfontein may also lead to the detection of traces of herbicides due to urban stormwater runoff. In other studies, agricultural runoff and urban surface runoff were also identified as the potential sources of pesticides in natural water sources [30,87,88]. Moreover, it was also observed during field sampling that some farmhouses had sceptic tanks. Rivers receive effluent from wastewater treatment works, are surrounded by activities such as animal husbandry, and are polluted by household wastes from nearby settlements. Therefore, it is reasonable to link emerging contaminants such as acetaminophen, ibuprofen, carbamazepine, estradiol, triclosan, progesterone, testosterone, and 17-alpha-ethinyl-estradiol to leakage from farmhouse sceptic tanks, discharge of wastewater effluents in streams, runoff from animal husbandry, and illegal dumping of household waste (containing unused/expired drugs) near the streams. In many parts of the world, human activities such as crop production, animal husbandry, illegal dumping, and wastewater effluents have been pinpointed as the main potential sources of many organic chemical pollutants in water sources [37,86,89,90,91].

Identification of Pollution Sources in Dams

The hierarchical cluster analysis in dams produced a dendrogram with four clusters in all seasons, as highlighted in Figure 3 and presented in Supplementary Table S11. In summer, C1 contained two contaminants (simazine and carbamazepine), which may be associated with wastewater effluent, illegal dumping, agricultural runoff, and domestic sewage. C2 contained seven contaminants (acetaminophen, progesterone, testosterone, estradiol, metolachlor, atrazine, and terbuthylazine), likely emanating from wastewater effluent, illegal dumping, domestic sewage, agricultural runoff, and urban surface runoff. C3 had two contaminants (triclosan and ibuprofen), which may be associated with wastewater effluent, illegal dumping, and domestic sewage. C4 contained one contaminant (17-alpha-ethinyl-estradiol), which may be associated with wastewater effluent, illegal dumping, and domestic sewage. During the autumn season, C1 contained nine contaminants (acetaminophen, testosterone, progesterone, atrazine, terbuthylazine, estradiol, metolachlor, ibuprofen, and carbamazepine), which may be connected to wastewater effluent, illegal dumping, domestic sewage, and agricultural runoff. C2 hosted one contaminant (simazine), most likely stemming from wastewater effluent and agricultural runoff. C3 had one contaminant (triclosan), possibly stemming from wastewater effluent, while C4 also had one contaminant (17-alpha-ethinyl-estradiol), possibly stemming from wastewater effluent, illegal dumping, and domestic sewage (Figure 3).

In the Modder River catchment, agricultural activities, namely crop production and animal husbandry, are the most prominent economic activity. Therefore, agricultural runoff may be a contributing factor to the occurrence of some emerging contaminants in surface water such as dams. Yang et al. [87] and Hou et al. [88] reported agricultural runoff as one of the potential sources of organic contaminants in surface water sources. Illegal dumping of domestic wastes (unused or expired drugs) and direct urination may also introduce traces of emerging contaminants into local reservoirs/dams as some of them are open to the public for picnics, aquatic sports activities, and conference centres events. Illegal dumping of domestic waste was also cited by other authors as one of the potential sources of chemical water pollution [86,91]. There was sewage runoff from a nearby settlement flowing into one of the dams during the field work; this may contribute to traces of some pharmaceuticals, including steroid hormones and personal care products. Many studies have also reported urban sewage and domestic wastewater discharge as potential sources of many chemical pollutants in water sources [86,89,90]. According to Gosset et al. [92], emerging contaminants are continuously released in trace amounts into receiving streams due to the inefficiency of wastewater treatment works seeking to remove them. Considering this case, these contaminants end up being introduced into local reservoirs/dams during the recharging period.

Identification of Pollution Sources in Treated Drinking Water

The hierarchical cluster analysis in treated drinking water also generated four clusters in all seasons, as highlighted in Figure 4 and presented in Supplementary Table S11. In summer, C1 hosted six contaminants (acetaminophen, progesterone, testosterone, estradiol, carbamazepine, and simazine), C2 had three contaminants (metolachlor, atrazine, and terbuthylazine), C3 had two contaminants (triclosan and ibuprofen), while C4 contained one contaminant (17-alpha-ethinyl-estradiol). During the autumn season, C1 was characterised by seven contaminants (acetaminophen, testosterone, progesterone, simazine, estradiol, carbamazepine, and terbuthylazine), C2 contained two contaminants (atrazine and metolachlor), C3 was the host of two contaminants (ibuprofen and triclosan), and C4 was characterised by one contaminant (17-alpha-ethinyl-estradiol), as shown in Figure 4. It is reasonable to conclude that dams or reservoirs used as a source of water for water treatment plants are polluted by wastewater effluent, agricultural runoff, illegal dumping, urban surface runoff, and domestic sewage. Wastewater effluents may introduce compounds such as acetaminophen, ibuprofen, progesterone, testosterone, estradiol, terbuthylazine, triclosan, atrazine, metolachlor, 17-alpha-ethinyl-estradiol, carbamazepine, and simazine into surface water. Agricultural runoff may carry traces of simazine, atrazine, metolachlor, terbuthylazine, and steroid hormones into surrounding water sources. Illegal dumping of expired or unwanted drugs may also pollute water resources. Domestic sewage may carry large amounts of contaminants such as acetaminophen, ibuprofen, progesterone, testosterone, triclosan, estradiol, terbuthylazine, atrazine, metolachlor, 17-alpha-ethinyl-estradiol, carbamazepine, and simazine. Urban stormwater may carry herbicides used in residential, commercial, and industrial areas as well as medical drugs dumped illegally on the streets into nearby water sources. Therefore, the contamination of rivers and reservoirs used as a source of water for water treatment plants may lead to the detection of these emerging contaminants in treated drinking water. Mugudamani et al. [30] reported that wastewater effluents and agricultural runoff in dams used as a source of water for water treatment plants may eventually lead to the detection of traces of emerging contaminants in treated drinking water. Therefore, water treatment works managers should continuously monitor their reservoirs in order to implement advanced treatment methods that can remove the detected emerging contaminants.

3.4. Interventions to Inhibit Water Pollution for Environmental Protection and Sustainability

In order to inhibit or reduce the load of emerging contaminants pollution in water sources for environmental protection and sustainability of water resources within the Modder River catchment, various types of interventions are necessary. Based on the source apportionment results in this study, emerging contaminants in the Modder River catchment are generally released mainly as a result of anthropogenic activities. The established potential anthropogenic sources of pollution include wastewater effluents, domestic sewage, urban surface runoff, agricultural runoff, and illegal dumping. Therefore, the following mitigation measures are suggested to help manage or remedy the situation.

3.4.1. Wastewater Effluents Discharged in Nearby Streams

Most wastewater effluents in the study area are discharged into nearby streams, thus introducing traces of emerging contaminants in those water bodies. According to Wang et al. [4], conventional wastewater treatment plants were not constructed to eliminate emerging contaminants. The inability of wastewater treatment works to remove emerging contaminants makes them a major point of release of emerging contaminants into the water environment. Therefore, relevant regulatory authorities should introduce proactive measures in the form of mandatory orders and formulate proper guidelines for permissible limits for the discharge of effluents in streams to reduce emerging contaminants pollution [93]. Considering that wastewater treatment works in this study receive industrial, domestic, and hospital wastewater together, there should be implementation and application of monitoring schemes that are able to discriminate between the domestic, industrial, and hospital wastewater in the area. Governments should also financially support current wastewater treatment works to develop and implement advanced treatment technologies (such as powdered activated carbon treatment, ozonation, peroxonation, etc.) for the removal of a broad spectrum of emerging contaminants with different properties [94]. Moreover, wastewater treatment works operators or managers should assess the overall mixture toxicity of emerging contaminants in effluents before discharge in order to protect aquatic life [95].

3.4.2. Illegal Dumping of Domestic Waste near Water Bodies

During the sampling campaign, illegal dumping of domestic waste (likely to contain unused or expired medical dugs) near some of the dams and rivers was witnessed. This situation may be due to lack of knowledge about the impact that organic contaminants may have on the environment. Moreover, lack of proper disposal means may also have intensified illegal dumping around the area. Therefore, educating people to segregate waste at the source and providing incentives to facilitate waste segregation may reduce illegal dumping in the area [96]. Taking back unused or expired medications may help to mitigate their quantities in the aqueous environment and their possible health risks. However, the effectiveness of this strategy rest on teaching people about the possible environmental implications of unused or expired medical drugs [93]. Therefore, local communities should be educated on the proper waste disposal methods and the environmental implication of illegal dumping.

3.4.3. Emission from Domestic Sewage Overflow

Domestic sewage overflow was considered to be one of the major sources for emerging contaminants emissions into water sources. From the field observation, some of the raw domestic sewage water overflows were channelled towards surface water. This situation means that majority of contaminants enter water resources without necessary treatment. Considering the increase in intensity and frequency of rainfall as a result of climate change, technical measures in the design of the sewer system are required in this area in order to reduce emission from these sources [95].

3.4.4. Emission from Urban Surface Runoff

Stormwater is a cause of concern in urban areas because urban areas are more polluted by a cocktail of anthropogenic emissions. The presence of emerging contaminants in stormwater may be a result of runoff, illegal dumping of wastewater and unsealed systems, and overflow in combined sewer systems. Moreover, the pollution of surface runoff by emerging contaminants is related to various activities such as the use of solid sludge from wastewater treatment works for soil amendments; the use of reclaimed wastewater for soil irrigation; the use of slurry and liquid manure as fertilizers; and urination and defecation of grazing animals and pets. Rainwater may carry pesticides from golf courses, parks, and residential properties through storm drains and then into rivers and local water reservoirs. Priority should also be placed on enhancing non-point source pollution control measures in the Modder River catchment, including the implementation of infrastructure improvements such as rainwater and sewage separation systems. These measures can effectively impede land-based emerging contaminants entering into water sources in the Modder River catchment [93].

3.4.5. Emission from Agricultural Activities

The Modder River catchment is surrounded by various agricultural activities such as crop production and animal husbandry. Crop production may necessitate the use of herbicides (atrazine, simazine, metolachlor etc.) for the maintenance of various crops while animal husbandry may require the use of medical drugs (fenbendazole, carprofen, estradiol, etc.) for the health of livestock. Emissions from these activities may introduce these compounds into nearby water bodies. According to Gozzo et al. [94], one of the main causes of pharmaceutical pollution is intensive farming, excretion through faeces and urine during the free grazing of animals, manure spreading on land, and contamination via runoff. Therefore, reducing both intensive farming and the use of livestock drugs could be crucial to guarantee the quality of surface waters within the Modder River catchment. Furthermore, improvements in agricultural practices may protect water bodies and contribute to the reduction of risks associated with the presence of emerging contaminants. Changes in treatment timings and intensities, as well as manure application rates and timings, may help to reduce discharges of emerging contaminants in the Modder River catchment [97].

Adoption of these proposed strategies will not only serve as a way to inhibit the pollution of river and local water reservoirs (dams) by emerging contaminants but are also a feasible means toward ensuring safer drinking water within the Modder River catchment. Moreover, protecting raw water from emerging contaminants pollution is an indicative premise of the water environment and ecosystem health. With the involvement of various stakeholders, the above mitigation measures may be essential for sustainable management of the Modder River catchment and the protection of public health.

4. Conclusions

Examining the occurrence and potential sources of emerging contaminants for the first time in the Modder River catchment brings to an end the dearth of data on emerging contaminants pollution in the Free State province. The present study suggests that all the sampled water sources have some level of questionable drinking water quality and necessitate some amount of treatment to reduce the contamination before consumption. It has also revealed that the Modder River catchment is vulnerable to pollution from emerging contaminants as a result of anthropogenic activities such as wastewater effluents, domestic sewage, urban surface runoff, agricultural runoff, and illegal dumping. Among the emerging contaminants detected, 17-alpha-ethinyl-estradiol should be listed as priority pollutant in future pollution monitoring and waste management programmes within the Modder River catchment in the Free State province. The outcomes of this study may be relevant for the prioritization of hazardous substances in order to create suitable monitoring campaigns and any necessary countermeasures adopted for environmental protection and sustainability of water resources. This work can facilitate the need to develop regulations aimed at reducing the spread of emerging contaminants in the Modder River catchment and other parts of the country.

5. Limitation and Future Studies

Although this study closed an important data gap on emerging contaminants pollution and their possible sources in the Modder River catchment, it has some limitations. The method used for sample analysis was only tested for selectivity, linearity, and limit of quantification. The non-target qualitative analysis was not considered in this study. The study also investigated few groups of emerging contaminants in few samples. Risks associated with identified emerging contaminants were also not assessed. Moreover, groundwater and wastewater effluents as sources of water were not monitored. Therefore, in future studies, non-target screening of emerging contaminants should be taken into consideration as it can lead to the discovery of still other classes of compounds in this catchment. More substances should be targeted so that a prioritisation of contaminants becomes feasible. Wastewater effluents and groundwater should also be analysed within the Modder River catchment. Additionally, assessing the individual and mixed ecological risks of emerging contaminants should be prioritised. This will help to reduce pollution for environmental protection and water sustainability within the Modder River catchment.