Reconnaissance Survey of Organic Contaminants of Emerging Concern in the Kabul and Swat Rivers of Pakistan
Abstract
The Swat and Kabul rivers of northern Pakistan are within an important regional watershed that supports river-based livelihoods and is impacted by untreated effluent discharges and municipal solid waste. Evidence indicates that fish populations are decreasing in these rivers. One potential cause of poor aquatic health is pollution; therefore, we investigated the presence of contaminants of emerging concern (CECs) in the river systems. Water samples were collected in the Kabul River (n = 9) and Swat River (n = 10) during seasons of high (summer 2018) and low (winter 2019) river flow. Agrochemicals, pharmaceuticals, plasticizers, chemicals in personal care products, and hormones were quantified via liquid chromatography high-resolution mass spectrometry. In the Swat River, caffeine (18–8452 ng/L), N,N-diethyl-meta-toluamide (DEET; 16–56 ng/L), and plasticizers (13–7379 ng/L) were detected at all sites during both seasons, while butachlor (16–98 ng/L) was detected only during high flow. In the Kabul River, caffeine (12–2081 ng/L) and several plasticizers (91–722 ng/L) were detected at all sites during both seasons, while DEET (up to 97 ng/L) was detected only during high flow. During low flow, pharmaceuticals (analgesics and nonsteroidal anti-inflammatory drugs) were quantified in both rivers (up to 823 ng/L), with detection frequencies from 70% to 100% and 0% to 78% in the Swat and Kabul Rivers, respectively. Intermittent-use and natural seasonal processes (increased runoff and dilution from rainfall and snowmelt) yielded higher agrochemical concentrations and lower concentrations of continuous-use compounds (e.g., caffeine) during high flow. The present study provides the first insight into CEC concentrations in the Swat River, additional insight into the Kabul River stressors, and, overall, contaminant risks to aquatic life. Environ Toxicol Chem 2023;42:2599–2613. © 2023 The Authors. Environmental Toxicology and Chemistry published by Wiley Periodicals LLC on behalf of SETAC.
INTRODUCTION
The term contaminants of emerging concern (CECs) refers to a variety of chemicals that are increasingly being detected at low concentrations in the environment, and thus are of increasing concern particularly in regard to aquatic life (Richardson & Kimura, 2020; Wilkinson et al., 2017). Contaminants of emerging concern include, but are not limited to, industrial chemicals, medications, food additives, personal care products and associated additives, perfumes, cosmetics, ultraviolet light filters, stimulants, and estrogenic compounds (Wilkinson et al., 2017). Endocrine-disrupting chemicals (EDCs), which are substances that hinder the normal operation of the endocrine system, consist of both CECs and legacy chemicals (Kabir et al., 2015; Lei et al., 2013; Matthiessen et al., 2018; Windsor et al., 2018; Writer et al., 2010). The presence and concentration of EDCs in aquatic ecosystems and freshwater resources continue to be of concern due to their detrimental health effects in humans (Kabir et al., 2015) and wildlife (Windsor et al., 2018), especially aquatic organisms like fish even at low concentrations (Lei et al., 2013; Matthiessen et al., 2018; Windsor et al., 2018; US Environmental Protection Agency, 2020). While several studies on CEC occurrence have been conducted, these have been predominantly in high-income countries (Lee et al., 2021), while water quality data in less developed countries were identified as a deficit in the World Water Development Report (Miletto et al., 2021).
In 2016, Pakistan became the first developing country to form a national development agenda toward accomplishing the Sustainable Development Goals; however, growing populations and water scarcity have critically impacted efforts to meet the performance goals for drinking water, hygiene, sanitation, solid waste management, and agriculture (Miletto et al., 2021; United Nations, 2022). For example, limited access to clean water has contributed to human health concerns and economic challenges in Khyber Pakhtunkhwa Province in northwestern Pakistan (Fida et al., 2022; Miletto et al., 2021). In particular, poor water quality has been cited as a contributing factor to regional fish declines in the Swat and Kabul rivers, which are essential water sources to Khyber Pakhtunkhwa Province (M. Ahmad et al., 2010; Hayat Khattak et al., 2015; Ishaq et al., 2014; Nafees et al., 2018; Yousafzai et al., 2015). These fish declines have been linked to adverse impacts to the livelihoods and businesses of locals (Mohammad et al., 2011; Nixon, Ma, Zanotti, et al., 2022). In a 2019 survey of Khyber Pakhtunkhwa household heads (n = 448), >90% of respondents reported observing decreased river water quality and 89% observed fish declines over the last 10 years (Nixon, Ma, Khan, et al., 2022). Even so, adequate monitoring of regional water quality is lacking.
The majority of contaminant studies in Pakistan have focused on metals and pesticides in portions of the Indus River basin downstream of Khyber Pakhtunkhwa in Sindh Province (Mirbahar et al., 2020; Scheurell et al., 2009) or Punjab Province (Ashfaq et al., 2016, 2019; Ashfaq, Khan, et al., 2017; Ashfaq, Noor, et al., 2017; Riaz et al., 2023). In the Khyber Pakhtunkhwa, prior contaminant studies focused primarily on metals in the Kabul (Ali & Khan, 2018; Aamir et al., 2017; K. Khan et al., 2022; M. I. Khan et al., 2018; T. Khan et al., 2011; Muhammad & Usman, 2022; Saad et al., 2017; Usman et al., 2017) and Swat rivers (Alam et al., 2020; Jehan et al., 2020; K. Khan et al., 2022; L. Khan et al., 2022; M. Khan et al., 2022; M. Liu et al., 2020; Muhammad & Usman, 2022; Sohail et al., 2022). Only a few studies in Pakistan have reported the occurrence of CECs in freshwater reservoirs, river water, and wastewater discharges in portions of the Indus River basin downstream of Khyber Pakhtunkhwa in Sindh Province (Mirbahar et al., 2020; Scheurell et al., 2009) or Punjab Province (Ashfaq et al., 2016, 2019; Ashfaq, Khan, et al., 2017; Ashfaq, Noor, et al., 2017; Riaz et al., 2023). Ashfaq et al. (2019) investigated 52 pharmaceutical and personal care products in urban drain wastewater and canal surface water during October 2015 in Lahore, Pakistan. Nonsteroidal anti-inflammatory drugs (NSAIDs) including ibuprofen, naproxen, and diclofenac (DCF) were reported up to 450 ng/L in Lahore surface water and 18 000 ng/L in wastewater with concentrations of both pharmaceuticals and chemicals from personal care products generally correlated with population density (Ashfaq et al., 2019). Wastewater and solid waste from pharmaceutical manufacturing in Lahore contained several hundred micrograms per liter to several thousand micrograms per kilogram of DCF, respectively (Ashfaq, Khan, et al., 2017). M. I. Khan et al. (2018) investigated eight pharmaceuticals (four NSAIDs and four antidepressant-related drugs) in Mardan's wastewater and river water during October to November 2016. They found paracetamol (acetaminophen) and ibuprofen at up to 104 and 102 ng/L, respectively, in the rivers, which were deemed to be of medium to high ecological concern and linked to open waste disposal. Of the studies conducted in Pakistan, only the latter one included the Kabul River and none included the Swat River. Riaz et al. (2023) were the first to report the occurrence of per- and polyfluoroalkyl substances (PFAS) in Pakistan. They surveyed PFAS (n = 54) in five freshwater reservoirs, with 15 PFAS frequently detected. More comprehensive occurrence studies on CECs are necessary to inform implementation of regional policies aimed at improving local water quality and aquatic ecosystem health in Khyber Pakhtunkhwa.
The present study was conducted as one component of a broader interdisciplinary collaboration designed to evaluate water quality–related concerns in Khyber Pakhtunkhwa Province. Specifically, a water sampling campaign was conducted along the Kabul and Swat rivers in the Upper Indus River basin with the goal of analyzing for an expanded list of emerging organic contaminants and evaluating seasonal impacts on contaminant occurrence and concentration.
MATERIALS AND METHODS
Study area
The Kabul River originates in Afghanistan and enters the territory of Pakistan through the Khyber District. The Kabul River, part of the greater Indus River watershed, is the chief source of water for aquaculture and agricultural irrigation in Khyber Pakhtunkhwa Province (Nafees et al., 2018). The Swat River, a major tributary of the Kabul River, is particularly critical for sustaining river-based tourism, irrigated agriculture, hydropower, fisheries, and fishing communities; however, it also receives pollution from industries, agriculture, and domestic sources (Mohammad, 2018; Saving river Swat from growing pollution, 2007). This region is vulnerable to climate change and flooding due to the strong seasonal water flow fluctuations from snowmelt and monsoons (I. Ahmad et al., 2015; Hussain & Mumtaz, 2014; Iqbal et al., 2018; Ullah et al., 2018), thereby increasing the vulnerability of this complex ecosystem to additional stressors resulting from river pollution. Baseflow contributions to the Swat River include groundwater, subsurface flow, and rain runoff (25%–50% of annual rain runoff), with increases in flow during the summer season due to runoff contributions from monsoon rains (30%–60% of annual rain runoff) and snowmelt (65%–75% of total runoff; Dahri et al., 2011).
Water in the Kabul River is routed into various channels known as the Shah Alam, Naguman, and Sardaryab (Adezai) rivers. The Kabul River flows through the Peshawar, Charsadda, and Nowshera districts and ultimately joins the Indus River near Attock (Figure 1). Contaminants of emerging concern enter the river system through direct discharge of untreated sewage or industrial effluent as well as several indirect discharges (Yousafzai et al., 2008). The regional Department of Environmental Planning and Management published a report that revealed many industries located in the city of Peshawar do not treat their wastewater prior to discharge into the Kabul River (Jers Engineering Consultants [JEC], 2017). Thus, open dumping of municipal solid waste (MSW) along with the release of untreated domestic and industrial wastewater effluents in Khyber Pakhtunkhwa Province have resulted in pollution of the rivers, irrigation networks, and available drinking water resources (W. Ahmad et al., 2021; Fida et al., 2022; Hussain & Mumtaz, 2014; Imran et al., 2018; Iqbal et al., 2022; H. K. Khan et al., 2022; T. Khan et al., 2011; Yousafzai et al., 2008). In some cases, excessive loads from industry in the Hayatabad Industrial Estate occur, as is the case for discharges coming into the Budhni stream that flows into the Kabul River (Figure 1). The Shah Alam River receives domestic wastewater from Peshawar through the Budhni canal as well as from areas surrounding Peshawar city (Nauthia Qadeem and Civil quarters through the Hazar Khuwani drain) and the Palosai area through the Hayatabd drain. Much of the industry in the Hayatabad Industrial Estate is food processing– and textile-related but also includes petroleum, rubber, plastic, and chemicals manufacturing; sugar mills; paper mills; mineral products; and metal products. Indirect discharges include stormwater runoff from different parts of the province (Swat, Charsadda, Mardan, Nowshera, and Peshawar), which is exacerbated by the open dumping of MSW, leading to both leachate from the dump piles as well as solid debris entering the rivers. In Mardan, MSW is utilized as a fill material to level residential lands (Israr et al., 2022). Butt et al. (2022) provides a comprehensive example of an unengineered dump site from 2019 located in the Peshawar basin.
Runoff from agricultural fields also occurs, which is the primary source of fertilizers and pesticides entering the Kabul and Swat rivers. In addition, most of the hospitals in the province have not been implementing the 2005 Hospital Waste Management Rules, which included segregation of solid risk waste (sharps, infectious waste, pharmaceutical waste, pathological waste, chemical waste, and radioactive waste), which are to be incinerated. Most hospitals do not have actively maintained incinerators; therefore, hospital solid waste as well as hotel solid waste are often combined with municipal waste dump piles (Khalid et al., 2021). Hospital effluents are also combined with sewage or drainage that is ultimately released into the river system (JEC, 2017; Khalid et al., 2021; Sadar et al., 2002). Hospitals, hotels, and health units along the Kabul and Swat rivers in the study sampling area along with industries are noted in Figure 1.
River sampling
The Kabul River system was sampled at nine sites (Figure 1 and Table 1), starting at the Warsak site (K1) sampling in a southeast trajectory following the Kabul River flow path to Jehangira (K9). The Swat River was sampled at 10 sites (Table 1), starting at Bahrain (S1) toward downstream Chakdara (S10). Different sampling approaches were used during high-flow (June 20–26, 2018) and low-flow (February 12–20, 2019) sampling campaigns. During the high-flow sampling, samples were collected using a solar-powered, battery-operated pump connected to tubing and a reusable solid-phase extraction (SPE) sorbent holder prototype loaded with fresh HLB sorbent disks (2 g, 47 mm; Atlantic HLB-H SPE disks; Horizon Technology) and glass fiber prefilters for each sample (Supporting Information, Figure S1). This approach was selected to improve sensitivity and facilitate processing large volumes of water in the field, thus also reducing processing time in the laboratory normally required for grab samples. For each sample, pump lines were purged with river water for approximately 5 min, disk holders were connected, and 3 L of river water were pulled through the extraction cartridge at 99.5 ± 3.3 mL/min, followed by drying the sorbent disk using a hand-operated vacuum pump. Cleaning of pump hoses between sample collections is detailed in the Supporting Information. Each set of sorbent pads and glass prefilters were packaged together in aluminum foil and placed in a labeled site-specific bag. During low river flow, 1-L grab samples were collected and passed through 200 mg Oasis® HLB SPE cartridges in a Supelco SPE system at the University of Peshawar Pakistan. Both the Atlantic HLB SPE disks from the high-flow sampling event and the Oasis HLB SPE cartridges from the low-flow event were shipped to Purdue University (West Lafayette, IN, USA) and stored at −14 °C prior to extraction and analysis of the CECs.
| Site | ID | Site specification |
|---|---|---|
| Warsak | K1 | Control point (upstream) |
| Shah Alam | K2 | Industrial wastewater from Hayatabad Industrial Estate Peshawar through Budhni Canal joins Shah Alam River at Hindko Daman Village, Peshawar. Shahi Katha drain passes through Nauthia and Civil Quarters of Peshawar, carrying all the domestic waste through Budhni Canal and meets Shah Alam River—a tributary of Kabul River. |
| Sardaryab | K3 | Sardaryab River passes through district Charsadda, which is a recreational point. Surrounding area is mostly associated with farming/agriculture. Swat River meets Sardaryab at Charsadda district. |
| Akbar Pura-I | K4 | Waste from Shah Alam River, Sardaryab River, and with addition of Swat River joins Kabul River at district Nowshera (Akbar Pura-I). |
| Akbar Pura-II | K5 | Bara River carries waste from the Khyber Agency and from the suburb of district Peshawar, which is dumped into Bara River through Hazar Khawani drain. Bara Canal further meets Kabul River at district Nowshera (Akbar Pura-II). |
| Amangarh | K6 | Industrial wastewater from Amangargh is released just upgradient of this point. |
| Akram Abad | K7 | Municipal waste of district Nowshera from surrounding villages is dumped into Kabul River in Akram. |
| Pir Sabaq | K8 | All the domestic and hospital waste of district Mardan flows through Kalpani River/Canal and is dumped at Pir Sabaq. |
| Jehangira | K9 | Downstream of Kabul River near Fauji Corn Complex—receives waste from surrounding villages. Afterward, near Khairabad, Kabul River meets Indus River. |
| Bahrain | S1 | Tourist spot upstream from all sampling sites, mostly occupied by hotels. |
| Madyan | S2 | Downstream to Bahrain, Madyan also has a number of hotels and trout fish farms. |
| Bagh Deri | S3 | Downstream of Kalakot stream, where domestic wastewater mixes with Swat River. |
| Khwaza Khela | S4 | Domestic wastewater of surrounding areas also mixes here with Swat River. |
| Manglawar | S5 | Manglawar stream joins Swat River here, which carries domestic wastewater along with hotel waste. |
| Sherabad (Takhtaband) | S6 | Water & Sanitation Services Mingora Swat dump all the solid waste of Mingora city at Sherabad in the form of dump piles along the riverbank. |
| Panjigram | S7 | All wastewater of Mingora city from cosmetic industry, plastic factories, hospitals, hotels, and domestic wastewater joins Swat River at Panjigram. |
| Barikot | S8 | It is downstream to Barikot Hospital. |
| Landakay | S9 | Domestic wastewater of surrounding areas joins Swat River here. |
| Chakdara | S10 | Downstream to all hotels, hospitals, factories, and domestic wastewater. The last sampling site of Swat River. |
Water velocity and general water parameters
River velocity was estimated at each sampling point using a float method, which requires observing the time for a floating object to traverse a known length and noting its position in the channel (Sanders, 1998). Ideally, for wide streams, aggregation of multiple measurements should be done at different depths and across the width of the river for an accurate measurement (Joy & Wierenga, 2005), but this was not possible in our study. For each sampling campaign, turbidity using a Micro TPW Turbidimeter field portable meter (HF Scientific) and dissolved oxygen (only 2019) using a Hanna DO Meter HI 9147 were measured on site. Separate grab samples were collected and brought to the Peshawar laboratory, where pH and electrical conductivity (microsiemens) were measured using an EZDO 7200 pH/conductivity meter.
Reagents
Six standard mixes were obtained (detailed in the Supporting Information) containing compounds for the most we categorize broadly as CECs or pesticides. In addition, isotopically mass-labeled compounds for diethyl phthalate (DEP), atrazine, triclosan, carbamazepine, caffeine, bisphenol A (BPA), and ibuprofen were obtained for evaluating and comparing extraction recoveries. Mass-label sources along with other reagents are detailed in the Supporting Information.
Sorbent disks and SPE procedures
Using a Horizon SPE system, the loaded Atlantic HLB-H sorbent disks with the glass prefilter from the high–river flow sampling were extracted sequentially using three sets of solvent schemes, twice with MeOH and acidified MeOH in the first two schemes, followed by acetonitrile into 40-mL glass vials (Supporting Information, Table S1) based on a US Geological Survey method validated for 500-mg Water Oasis HLB SPE cartridges (Furlong et al., 2008). Oasis HLB SPE cartridges (Atlantic HLB SPE) from the grab samples were extracted using a Supelco SPE system in a similar manner as the disks, except two additions of methanol were followed by acidified methanol and samples were allowed to elute by gravity into glass centrifuge tubes. For both disks and cartridges, extracts were combined, evaporated to approximately 0.5 mL, and quantitatively transferred to preweighed injection vials. Seven isotopically mass-labeled compounds were added as internal standards (ISs) to injection vials, and the final injection solution composition was 40:60 v:v MeOH:H2O. Additional details are in the Supporting Information.
Organic compound analysis
Organic compound analysis was conducted using a Shimadzu ultrahigh-performance liquid chromatography (uPLC) system coupled to an AB Sciex 5600 quadrupole time of flight mass spectrometer (QToF-MS) operated in both negative and positive electrospray ionization (ESI) modes. General uPLC/QToF-MS system operating parameters are detailed in Supporting Information, Table S2, and compounds targeted in negative and positive ESI mode are listed in Supporting Information, Tables S3 and S4, respectively. Samples were injected (20 µL) onto a Phenomenex Kinetex Evo C18 column and eluted at a flow rate of 0.5 mL/min with a binary gradient consisting of 0.15% acetic acid in H2O and MeOH for negative ESI analysis and 10 mM formate buffer solution (ammonium formate and formic acid, pH 3.5) in H2O and MeOH for positive ESI analysis (Supporting Information, Table S2). Data acquisition and quantification were performed using Sciex Analyst TF1.7 and Sciex Multiquant 3.0.1 software, respectively. CEC concentrations were calculated using a five- to seven-point calibration curve, generally ranging from approximately 10 to 500 µg/L. For the seven compounds for which there were ISs, concentrations were calculated using isotope dilution.
Quality control and method validation
All procedures for method validations, quality control, limit of detection (LOD), limit of quantitation (LOQ), and method LOQ (MLOQ) are detailed in the Supporting Information and Table S5. All sample processing was tracked gravimetrically. Solvent blanks consisting of the solvents used in each of the SPE dilution schemes and H2O blanks were processed identically to the samples. Recovery was assessed by processing spiked deionized water (H2O) samples through the appropriate extraction process. Target concentrations for the 1-L grab samples processed through the 200-mg Oasis HLB SPE cartridges were 100 and 500 ng/L and for the 3-L samples processed through the Atlantic 2-g HLB SPE disks, 100 ng/L. For the high-flow sampling system, recovery tests and contaminant breakthrough were evaluated simultaneously using two disks in series followed by a contaminant carryover test. The percent recovery was evaluated by comparing compound mass recovered in the first disk for 3-L spiked deionized H2O samples (n = 3) to spike checks (spiked directly into an injection vial). Contaminant breakthrough was assessed by comparing the relative mass recovered in each of the two disks in series. Immediately after the recovery tests, deionized H2O was pumped through the system for 5 min, followed by pumping 3 L of H2O (Process Blank) through the system, as was done with river water at each sampling site to evaluate sample carryover.
RESULTS AND DISCUSSION
Quality control and method validation
Matrix effects
For the seven compounds that had paired mass-labeled ISs (three for negative ESI and four for positive ESI), concentrations estimated with and without the IS corrections did not vary significantly (p > 0.5) based on a Wilcoxon test (Supporting Information, Figure S2) for process control samples spiked at the upper end of the calibration curve where suppression or enhancement is expected to be the greatest.
Recoveries
Compound recoveries from the 1-L grab samples ranged from 9% to 158% across all compounds (average of 93 ± 33%, median of 102%), with 75% of the compound recoveries in the 70% to 130% range (Supporting Information, Figure S3), generally consistent with the literature (Supporting Information, Table S6). For the high-volume sampling system, the first sorbent disk captured >75% of the total mass captured (sum of both disks in series; Supporting Information, Figure S4); thus, contaminant breakthrough appeared limited. Additional discussion and details are provided in the Supporting Information. Carryover in the high-volume system tubing was <2% relative to the spiked water sample processed immediately before the process blank (with a 5-min flush in between as performed in the field) and near MLOQs. In the carryover test, tributyl phosphate was identified as a system contaminant; tributyl phosphate was not one of the compounds in our mixtures used to assess recoveries.
Water quality and river flow
Water quality parameters measured in high- and low-flow seasons in both the Kabul and Swat rivers are summarized in Supporting Information, Tables S7 and S8, respectively. Estimated river flow ranges during low- and high-flow seasons for the Kabul River were 0.35 to 3.7 and 1.61 to 5.51 km/h. The mountainous Swat River showed higher river flow velocity values than the Kabul River at both low flow (0.35–6.85 km/h) and high flow (4.38–9.34 km/h). The Kabul River flow during low flow was highest (3.7–6.59 km/h) at the farthest upstream site (Warsak, K1). During high flow, although the Warsak site (K1) fell among the higher flows, the highest flows were at the points where other streams converged into the Kabul, for example, K5 where the Bara River and the Bhundi stream enter the Kabul as well as after the Kalapani stream enters the Kabul (K8 and K9). For the Swat River, the northernmost sampling site (S1) had higher flows during both low- and high-flow seasons. Turbidity increased during the high-flow seasons as expected. The Kabul River generally was more turbid than the Swat for both seasons but particularly during high flow for which turbidity exceeded 240 nephelometric turbidity units at all sites except the one on the Swat (K3) immediately prior to entering the Kabul River. Dissolved oxygen was not limiting with regard to aquatic health in both rivers and seasons (>7.9 mg/L), and pH was generally in the 7.5 to 8.4 range, which is typical for rivers. Electrical conductivity ranged from 100 to 1000 μS, with the Kabul River consistently having higher conductivities than the Swat and conductivity generally being higher during the high-flow season. The general water quality parameters measured align well with those reported for northwest Pakistan (Shah et al., 2019).
Organic chemical occurrence
Of the 67 compounds targeted for analysis, only 18 were found in one or more of the 2018 or 2019 samples (Table 2). The data reported for the samples were not corrected for recovery because of significant variability in concentration-specific recoveries and water volume relative to sorbent, as previously demonstrated (Furlong et al., 2008; Grujić et al., 2009), as well as similarities reported across sampling methods in the recoveries for many compounds in the present study. The detection frequency and concentration ranges for quantified compounds are included in Table 2 along with compound abbreviations. Compounds detected in the Kabul River included five pharmaceuticals, six phthalates, two pesticides, caffeine, and BPA (Supporting Information, Table S9). Compounds detected in the Swat River included six pharmaceuticals, eight phthalates, two pesticides, caffeine, and BPA (Supporting Information, Table S10). In general, most CECs were observed at higher concentrations in the samples collected during the low–river flow sampling event (February 2019). As previously mentioned, additional water contributions from snowmelt and monsoon rains to both rivers drive high-flow conditions and often lead to regional flooding. High-flow conditions lead to dilution of compounds from many sources in river water, particularly sources not directly affected by runoff (e.g., domestic waste). Seasonal runoff during high flow, as described in the Swat River, consists of monsoon rains (up to 60% of yearly rain runoff) and snowmelt (up to 75% of total runoff; Dahri et al., 2011) and can offset dilution by introducing CECs to the rivers from sources like agricultural fields and MSW dumping sites. Specific spatial and seasonal trends will be discussed within chemical classes.
| Swat River | Kabul River | |||||||
|---|---|---|---|---|---|---|---|---|
| 2018 | 2019 | 2018 | 2019 | |||||
| Range (ng/L) | Detection frequency (%) | Range (ng/L) | Detection frequency (%) | Range (ng/L) | Detection frequency (%) | Range (ng/L) | Detection frequency (%) | |
| Pharmaceuticals | ||||||||
| Acetaminophen | ND–LOQ | 20 | 83–612 | 70 | ND | 0 | ND–144 | 11 |
| Naproxen | ND–LOQ | 20 | 14–207 | 100 | ND | 0 | ND–112 | 78 |
| Sulfamethoxazole | LOQ | 100 | LOQ–41 | 100 | ND | 0 | ND–76 | 33 |
| Ibuprofen | ND–21 | 40 | ND – 379 | 80 | ND–11 | 11 | ND–416 | 78 |
| Diclofenac | ND–LOQ | 30 | ND – 246 | 90 | ND | 0 | ND–520 | 78 |
| Trimethoprim | ND–LOQ | 10 | ND–LOQ | 70 | ND | 0 | ND | 0 |
| Sum | LOQ–21 | – | LOQ–823 | – | ND–11 | – | ND–762 | – |
| Agrochemicals | ||||||||
| Butachlor | –LOQ–98 | 100 | ND | 0 | ND–17 | 11 | ND | 0 |
| N,N-Diethyl-meta-toluamide | LOQ–23 | 100 | –LOQ–56 | 100 | 18–55 | 100 | –ND–97 | 70 |
| Plasticizers | ||||||||
| Diethylphthalate | ND | 0 | ND–47 | 60 | 59–95 | 100 | 46–283 | 100 |
| Σ Dibutyl phthalate | ––LOQ–49 | 100 | 24–663a | 100 | LOQ–17 | 100 | 14–116 | 100 |
| Benzyl butyl phthalate | LOQ–58 | 100 | LOQ–18 | 100 | LOQ–17 | 100 | LOQ | 100 |
| Bis (2-ethylhexyl) phthalate | 13–51 | 100 | 179–6667 | 100 | 25–71 | 100 | 104–314 | 100 |
| Dinonyl phthalate | LOQ–33 | 100 | 57–345 | 100 | ND–44 | 56 | 49–95 | 100 |
| Dimethyl phthalate | LOQ | 100 | –ND–LOQ | 40 | ND | 0 | ND | 0 |
| Bisphenol A | LOQ | 100 | ND–86 | 50 | ND | 0 | 7–49 | 100 |
| Sum | 13–176 | – | 328–7379 | – | 91–260 | – | 240–722 | – |
| Additional CECs | ||||||||
| Caffeine | 18–129 | 100 | 93–8452 | 100 | 12–105 | 100 | 251–2081 | 100 |
- a One S6 replicate had a very high concentration of dibutyl phthalate (25 237 ng/L) and no other compounds, so was excluded from the reported average and range.
- ND = not detected; LOQ = limit of quantitation; CECs = contaminants of emerging concern.
Caffeine
Caffeine, while a well-known stimulant found in black tea and coffee, is considered a tracer of human activity; but it is also often combined with pharmaceuticals and thus could also be associated with direct discharges from pharmaceutical companies. Caffeine concentrations at each sampling site are reported in Figure 2. Caffeine was present at all sampling locations, with concentrations in the Kabul River higher during low flow (March 2019, 102–103 ng/L) compared to high flow (June 2018, 10–102 ng/L). Similar trends and concentration ranges were present for caffeine in the Swat River. Caffeine recoveries were comparable between sampling systems; thus, it is reasonable to assume the lower concentrations under high-flow seasons are primarily due to dilution of wastewater discharge from hotels during the summer tourism season because domestic discharge can be assumed to be constant.
Overall, caffeine concentrations observed in the Kabul and Swat rivers are in the range observed in Lahore, Pakistan (population >11 million; located in Punjab Province neighboring Khyber Pakhtunkhwa Province), with a mean concentration of 74 ng/L in the Lahore Canal up to a range of 4560 to 6940 ng/L in the Cantonment and Sham wastewater drains (Ashfaq et al., 2019). During low flow, the influence of continuous inputs of caffeine from wastewater in the Kabul River is observed in the Shah Alam tributary (K2) and in subsequent downstream samples (K4–K9), where concentrations were a magnitude higher than upstream samples (2081 and >1300 ng/L, respectively). In the Swat River similar downstream increases are observed near population centers during low flow; however, domestic wastewater inputs from seasonal tourism events led to upstream shifts for areas with high caffeine concentrations from downstream population centers (S7, S9) to upstream hotel clusters (S5) during high flow.
Pharmaceuticals
Pharmaceutical contamination in the region originates from untreated waste from pharmaceutical industries, hospitals, and domestic sources (Figure 2). In the Kabul and Swat rivers, CECs from three major classes of pharmaceuticals were detected including analgesics (acetaminophen, also known as paracetamol), NSAIDs (naproxen, ibuprofen, diclofenac), and antibiotics (sulfamethoxazole; Table 2). In the Swat River, trimethoprim (antibiotic) was also detected but below the LOQ. Low recoveries of many of the pharmaceuticals (e.g., erythromycin and ciprofloxacin) likely contributed to the limited number detected.
Similar to caffeine, pharmaceutical concentrations were typically higher during low flow, with the site-specific average sum (∑) of five pharmaceuticals across sampling sites along both rivers ranging from not detected (ND) to approaching 103 ng/L, whereas only ibuprofen was >LOQ (up to 21 ng/L) during high flow (Figure 3). In the Kabul River, sulfamethoxazole, naproxen, ibuprofen, and DCF concentrations ranged from ND to 520 ng/L, with DCF being the highest. Diclofenac in regional surface water is a concern because it has been linked to renal failure and population decline in Gyps bengalensis, an endangered vulture species (Oaks et al., 2004). Upper concentrations of ibuprofen and DCF in the present study were tens to thousands of times higher than those previously reported for the Kabul River (A. Khan et al., 2018). Common open dumping of MSW in Hayatabad Industrial Estate located in Peshawar (Sadar et al., 2002) and along the Bara River and the Budhni Canal as well as wastewater from industry are identified as sources of pharmaceuticals to the Kabul River. Acetaminophen was only detected in one sample along the Kabul River (144 ng/L, K2) at the site impacted by the Hayatabad Industrial Estate in the Peshawar District (Figure 1), which includes an acetaminophen manufacturing facility.
For the Swat River, the detection frequency (df; percentage) was lower during high flow compared to low flow for acetaminophen (20% and 70% df, respectively), naproxen (20% and 100% df), ibuprofen (40% and 80% df), DCF (30% and 90% df), and trimethoprim (10% and 70% df). The highest concentrations occurred downstream of the Mingora population center and were heavily influenced by domestic and hospital wastes, with sulfamethoxazole, naproxen, ibuprofen, and DCF ranging from LOQ–41 ng/L, LOQ–207 ng/L, ND–379 ng/L, and ND–246 ng/L, respectively.
Agrochemicals and insecticides
Agriculture is the backbone of Pakistan, a cornerstone of the local economy, particularly in Khyber Pakhtunkhwa because of the fertile soil of the province and availability of freshwater resources for irrigation. Pesticides are used in orchards, cereals, rice, and sugarcane production, and thus can enter rivers through agricultural runoff. The illegal use of pesticides (e.g., endosulfan and cypermethrin) for fishing has also been reported in Khyber Pakhtunkhwa (Mohammad et al., 2011). Agrochemicals do not follow the same seasonal trends as pharmaceuticals or caffeine because their immediate use during the growing season has greater impacts on their concentrations in river water than the dilution that occurs during high flow. In addition, high flow is the result of both snowmelt and monsoons that result in agrochemical-containing soil runoff.
Butachlor (BTC) is used as a pre-emergent in rice crops. Rice is a kharif (monsoon season) crop in Khyber Pakhtunkhwa and commonly grown along the Swat River. Butachlor was detected at all sites during high flow (summer) in the Swat River and was not detected during low flow or in the Kabul River itself. However, BTC was reported in the downstream Swat River site that is grouped with the Kabul River system sampling locations. The highest concentration (98 ng/L) was reported at Panjigram, while concentrations were lower (17 ng/L) before the confluence of the Swat River with the Kabul River at Sadaryab, which is a recreational spot for fishing and swimming. Upstream sites on the Swat River were <LOD. In addition, many legacy pesticides used in the region were not included in the present study because they have been previously reported (Aamir et al., 2017; Saad et al., 2017). For example, endosulfan is a commonly used pesticide reported at concentrations up to 166 ng/L in the Indus River (Shaikh et al., 2014).
N,N-Diethyl-meta-toluamide (DEET) is an insect repellant, and most of its applications are for personal-care use. Mostly domestic use and disposal of waste containing DEET are contributing to sources in the present study. DEET was detected in all the Swat River sampling locations for both seasons, albeit in most cases <LOQs, while most Kabul River samples, DEET was >LOQ (83% df; Figure 4). The highest DEET concentrations in the Kabul River were similar for high- and low-flow conditions (55 and 97 ng/L, respectively) and likewise for the Swat River (23 and 56 ng/L for high- and low-flow, respectively). Highest DEET concentrations were downstream of population centers linking DEET to untreated wastewater discharges from various areas similar to that noted for pharmaceuticals. A nontarget study reported DEET within sewage drains was being releasing into the Indus River, Pakistan (Shaikh et al., 2014). It was detected ubiquitously in Singapore, with median concentrations of 1140 ng/L in raw wastewater, 285 ng/L in treated wastewater, 119 ng/L in urban stormwater runoff, 83 ng/L in agricultural stormwater runoff, and 60 ng/L in receiving surface waters (Tran et al., 2019). This implies that domestic effluent, runoff, and agricultural sources may contribute to the concentrations reported in the present study.
Plasticizers
Plasticizers are commonly used to add flexibility to polymer products and may compose up to 50% of polyvinylchloride (Petersen & Jensen, 2010). Many plasticizers are confirmed EDCs including BPA and phthalates. Industrial production of plastics, food packaging, and personal care products are common sources of phthalates. Plasticizers also leach out from the products during manufacturing and disposal stages (Katsikantami et al., 2016; Marttinen et al., 2003). Fikarová et al. (2019) observed in lab-based studies the rapid leaching of some plasticizers into seawater from polyethylene including BPA, dimethyl phthalate (DMP), and DEP, as well as DMP and DEP from polyvinyl chloride (PVC), which are characteristic of their relatively low octanol–water partition coefficients (KOW). Other plasticizers with higher KOW values, such as benzyl butyl phthalate (BBP) and dibutyl phthalate (DBP), leached more slowly but continually. The European Union has legislation banning phthalates >0.1% in food contact products, to prevent reproductive impacts from exposure (Petersen & Jensen, 2010). Cosmetics often include lower–molecular weight phthalates such as DMP used in hair spray and DEP used in fragrances (Bergé et al., 2013; US Food and Drug Administration [FDA], 2022; Zulfiqar et al., 2015). Dibutyl phthalate has a slightly larger mass and is used in nail polish, epoxy resins, and adhesives. Greater–molecular weight phthalates have different applications, like bis-(2-ethylhexyl) phthalate (BEHP) used in PVC products and BBP used in consumer goods and medical products (Bergé et al., 2013; USFDA, 2022). The regional sources for phthalates include industries and domestic use. Hayatabad Industrial Estate has some factories manufacturing PVC cables and other products. In addition, phthalates are used in cosmetics and different beauty products that are available in Pakistan. Plastics and food packaging may be a considerable source of phthalates and particularly BEHP (Erythropel et al., 2014).
Due to the ubiquity of phthalates, leaching was assessed to inform the interpretation of data (Supporting Information, Figure S4-A). Some leaching occurred from the high-volume sampling system releasing BBP (approximately 8% of the total spiked amount), which contributed to high recoveries. However, BBP was not detected at many sampling sites, and the amount leached from the tubing was only high enough to impact concentrations reported for sites S2, S3, and S8. The recovery for BBP with high volume sampling system is >200% and far exceeds the recovery of the grab samples, likely explaining the low detection during 2019. Some leaching of DBP also occurred, but the method reporting limit in samples was greater than the amount leached. In both water and methanol, BEHP had high leaching contributions, with methanol increasing the concentration leached to subsequent pore volumes measured over time and in water after exposure to methanol. Bis-(2-ethylhexyl) phthalate is ubiquitous and was reported to be present in food contact surfaces, gaskets, and single-use gloves above the regulatory limits in the European Union (Petersen & Jensen, 2010). Thus, it is likely that leaching of phthalate occurs from the high-volume sampling system (e.g., fluorinated ethylene propylene tubing or gaskets) in the presence of methanol in addition to desorption of previously sorbed compounds. However, the reported concentrations in samples were higher than the amount leached and should still be considered in the context of the present study.
Phthalates (DBP, BBP, BEHP, and dinonyl phthalate) were detected at almost all sampling locations in both rivers (Figure 5), but seasonal factors (e.g., dilution, summer tourism, and temperature) appeared to impact phthalate concentrations differently in each river system. In the Swat River DBP, BEHP, and dinonyl phthalate were reported at all locations during both low and high flow, while BBP, BPA, DMP, and dioctyl phthalate were detected at all locations during high flow and at 100%, 80%, 50%, and 40% df during low flow, respectively. In the Kabul River, dioctyl phthalate and DMP were not detected, while DEP, DBP, BEHP, and BPA were detected in all samples and dinonyl phthalate at 78% df. In almost half the samples, BBP was only detected during high flow (44% df). The higher frequency of detection in the Kabul River is most likely a result of population density and personal care products or cosmetics used year-round.
Contributions to DBP present in all samples likely include industrial effluents, domestic effluents, and MSW. Among all detected phthalates, the highest concentration was of BEHP; BEHP is commonly sourced from plastics, which are abundant due to a lack of MSW infrastructure and open dumping. In the Kabul River, during high flow (2018), the Akbar Pura-I site revealed the highest sum plasticizer load, ∑260 ng/L; but during low flow (2019), the Shah Alam River had the highest plasticizer load, ∑722 ng/L. In the Swat River, Sherabad had the highest concentration during low flow, ∑7379 ng/L, and Landakay had the highest concentration during high flow, ∑150 ng/L. Previously, DEP, DBP, and BEHP were reported in the Indus River system, with the highest concentrations up to 386, 400, and 1965 ng/L, respectively, reported for discharges from wastewater drains into the river (Shaikh et al., 2014).
Bisphenol A is an industrial chemical that is primarily used for manufacturing plastic. It is also used in the packing of certain food items in cans and bottles. The major source of exposure to BPA is food (Schecter et al., 2010). In the Swat River, BPA was detected at all locations during high flow; however, during low flow, BPA (ND–86 ng/L) was only detected downstream beginning at Manglawar. In the Kabul River, BPA was only detected during low flow, with a concentration detected at Akbar Pura-II of 49 ng/L. BPA sources in Akbar Pura-II may be the solid waste disposal site of Hazar Khawani. Leaching out of plastic from solid waste might have caused the high concentration of BPA at Akbar Pura-II because the Bara River passes through Hazar Khawani and meets Kabul River at Akbar Pura.
Plastics may be a considerable source of phthalates, particularly BEHP in the Kabul River because of leaching from improper disposal of municipal plastic wastes. Peshawar City lacks proper solid waste disposal practices (Manzoor et al., 2020). Solid waste is dumped openly at two sites off of the ring road, one at Hazar Khwani and the other at Pishtakhara (an area the Bara River passes). This may be one of the reasons for the higher phthalate concentration at Akbar Pura-I site. According to the latest Khyber Pakhtunkhwa Environmental Protection Agency report, the major hospitals of the province (i.e., Lady Reading Hospital, Khyber Teaching Hospital, and Hayatabad Medical Complex) produce 1036, 413, and 260 kg/day hospital waste, respectively (Asian Development Bank, 2022). The incinerators of those hospitals are not in good condition. In addition, approximately 602 tons/day of MSW are produced in Peshawar City, and most of this goes into the sewage pipes and drains. These conditions are worse in other cities of the province. For example, the Sherabad sampling location is near an MSW dump on the banks of the Swat River that receives waste from Mingora. This unmanaged waste stream can contribute phthalates through leaching from plastics or through contamination of ground and surface waters via landfill leachate. Landfill leachate is known to contain phthalates. In China, up to ∑139 000 ng/L phthalates (e.g., BEHP, DMP, and DBP) was found in leachates, while contaminated groundwater and surface water had up to ∑14 000 and ∑5100 ng/L, respectively (H. Liu et al., 2010). Near upstream sites on the Swat River, MSW is dumped directly into the river because of a lack of sanitation and solid waste management infrastructure, thus exacerbating impacts downstream prior to subsequent dilution from tributaries.
Risk to aquatic life
Risk to aquatic life of reported organics was evaluated by comparing the maximum reported concentration in each river to the no-observable-effect concentration (NOEC). A ratio of the environmental concentration to NOEC >1 was indicative of a possible risk to aquatic life (Table 3). In both river ecosystems, BEHP poses the highest risk to aquatic life, with concentrations in the Swat River during low flow posing the highest risk. During low flow, ibuprofen also poses a risk to aquatic life in both rivers. The toxicity values utilized for this assessment were obtained using typical fish species (e.g., zebrafish, Japanese medaka, fathead minnow) listed in Supporting Information, Table S11. Additional interpretation of these results in the region is limited by the availability of toxicity data on the fish species of interest.
| Compound name | Maximum concentration (ng/L) | Reported toxicity concentrationa (ng/L) | Risk characterization | Designation | ||
|---|---|---|---|---|---|---|
| Swat River | Kabul River | Swat River | Kabul River | |||
| Acetaminophen | 612 | 144 | 1000 | 610 | 140 | Continue monitoring |
| Ibuprofen | 379 | 416 | 100 | 3790 | 4160 | High risk |
| Caffeine | 8452 | 2081 | 120 000 | 70 | 20 | Low risk |
| Bis(2-ethylhexyl) phthalate | 6667 | 314 | 10 | 666 700 | 31 400 | Very high risk |
| Bisphenol A | 86 | 49 | 1000 | 90 | 50 | Low risk |
- a Toxicity concentrations for various fish species were obtained using the Hazard Comparison Dashboard, which reports values in micrograms per liter, but were converted to nanograms per liter, consistent with the monitoring data. Additional details are available in Supporting Information, Table S11.
- Risk was characterized at the highest detected concentrations relative to the concentrations reported to have no observable effect. Compounds determined to have a very low risk (ratio <0.01) are not included. Bold values are those considered to have the designation in the last column.
CONCLUSION AND RECOMMENDATIONS
The present study was especially challenging because of governmental restrictions surrounding travel and transport of samples; however, studies like this one are necessary in developing regions to inform policy and infrastructural investments by both local governments and entities that invest in sustainable development like the World Bank. The practicality of sample collection and transport can greatly constrain CEC research. In the present study, we used high-volume sampling in an effort to minimize these challenges and obtain more representative samples. This approach was especially useful in the dangerous high-flow conditions of the Swat and Kabul rivers, where grab samples could not be safely obtained. Moreover, the high-volume sampling resulted in easily transported samples that required less additional sample processing than the more conventional approach. These benefits should be considered for similar studies in poorly accessible areas.
Contaminants of emerging concern were quantified at concentrations that were lower than expected. Phthalates may require continued monitoring in the Kabul and Swat rivers. Possible sources of contamination were broadly identified, and further studies should isolate specific sources, particularly along the Bara River in the Kabul River and Mingora in the Swat River. In particular, the role of improper disposal of MSW should be addressed regionally to reduce phthalate loads. Dilution plays a significant role in decreasing the pollution load in both rivers because high concentrations of CECs were recorded for the low-flow (2019) season. Non-point-source pollution also plays a major role in increasing the pollution load. However, point-source pollution can be addressed by establishing water resource recovery facilities in the province. Likewise, the proper implementation of MSW management and infrastructure would greatly decrease regional contamination. Currently, the environmental quality standards of the Khyber Pakhtunkhwa Environmental Protection Agency do not include regulatory guidelines for CECs. Fida et al. (2022) further discuss water-based policy and the lack of regional resources for provincial governments to implement the recommendations mentioned in the national water quality standards. Moreover, 2025 was mentioned as the year when water quality benchmarks were to be met. The Environmental Protection Agency should develop standards for the regular monitoring of freshwater bodies across the province to preserve freshwater assets for aquatic health and recreational purposes. Monitoring should include compounds with identified risks like BEHP and ibuprofen.
Multiple water resource stressors in the province are also deteriorating the quality and quantity of the Kabul and Swat rivers and their valuable aquatic ecosystems. Currently, the Khyber Pakhtunkhwa River Protection Act, 2014, is in place; but additional regional investments and enforcement measures must be taken for its successful implementation to preserve the aquatic assets of the province. Thus, decision makers and stakeholders must invest in infrastructure and the successful implementation of regional regulations for river protection and local health.
Supporting Information
The Supporting Information is available on the Wiley Online Library at https://doi.org/10.1002/etc.5750.
Acknowledgments
This work was funded by a grant from the US government and the generous support of the American people through the US Department of State and the US Agency for International Development under the Pakistan–US Science & Technology Cooperation Program. The contents do not necessarily reflect the views of the US government. This project was also supported in part by the US Department of Agriculture, National Institute of Food and Agriculture Hatch Funds (accession no. 1006516), a Purdue Andrews PhD Fellowship, and a Purdue Bilsland Dissertation Fellowship.
Disclaimer
Not applicable other than what is in the acknowledgment as required by the agency that funded this work.
Author Contributions Statement
Maria Christina Schilling-Costello: Formal analysis; Investigation; Methodology; Writing—original draft. Neelam Asad, Peyman Yousefi: Investigation; Methodology; Writing—original draft. Muhammad Haris: Investigation. Bushra Khan: Conceptualization; Funding acquisition; Investigation; Methodology; Supervision; Writing—review & editing. Linda S. Lee: Conceptualization; Formal analysis; Funding acquisition; Investigation; Methodology; Project administration; Supervision; Writing—original draft; Writing—review & editing.
Open Research
Data Availability Statement
All data are available in the Supporting Information.
