Freely Dissolved Organochlorine Pesticides (OCPs) and Polychlorinated Biphenyls (PCBs) along the Indus River Pakistan: Spatial pattern, Air-water Exchange and Risk Assessment

Freely dissolved OCPs and PCBs were measured by using polyethylene passive samplers, at 15 sites during 2014 throughout the stretch of Indus River to investigate spatial pattern, air−water exchange gradients and risk assessment. Concentrations (pg/L) of dissolved ∑ OCPs and ∑ PCBs ranged from 34– 1600 and 3–230. Among the detected OCPs, dissolved DDTs (p,p′-DDE, followed by p,p′-DDT) predominated with levels of 0.48 to 220 pg/L. The order of occurrence for other studied OCPs was as follows; HCB, Endosulfans, Chlordanes, and HCHs. Spatially, dissolved (pg/L) ∑ OCPs varied (p < 0.05) as: surface water of Alluvial Riverine Zone (ARZ) showed highest levels (114) followed by Frozen Mountain Zone (FMZ) (52.9), Low Lying Zone (LLZ) (28.73) and Wet Mountain Zone (WMZ) (14.43) respectively. However, our zone wise PCBs data did not exhibit signicant differences (p > 0.05). The PCA/MLR results showed that pesticides usage in crop/orchard elds and health sector, electric & electronic materials, and widespread industrial activities as the main source of OCPs and PCBs along Indus River. Air-water exchange of OCPs at FMZ, o,p′-DDE, p,p′-DDE and o,p′-DDD exhibited net volatilization while p,p′-DDD/o,p′-DDT showed net deposition, while all other studied zones (i.e., WMZ, ARZ and LLZ) showed net deposition of DDTs. Nevertheless, air-water exchange of PCBs showed that, there was net volatilization at the FMZ, WMZ and ARZ and net deposition at LLZ. Our results showed that OCPs and PCBs contaminated water intake, played an important role towards the considerable cancer/non-cancer risk (HI and CR values) along the Indus River Flood-Plain.

OCPs and PCBs are also well known semi-volatile organic compounds (SVOCs), which have the potential to volatilize and re-emit from secondary sources (from previously contaminated environments), undergo long-term and short-term atmospheric transport (in semi-arid/arid areas), and ultimately deposited into remote colder and/or high altitudinal locations, in particular north, south and third pole environment (Nasir et al. 2014, Wania &Haugen 1999. The process of transport of such SVOCs to long distances mainly occurs due to temperature and altitudinal variation and is well known as "Long Range Atmospheric Transport (LRAT)" and con rmed by many studies (Cincinelli &Dickhut 2011, Daly &Wania 2005, Davidson et al. 2003). Moreover, studies have also con rmed that remote colder areas acts as repositories of these chemicals, and in context of climate change, these repositories become highly active for remobilizing the chemicals in air and to the freshwater resources, which resulted in a major threat to human and wild life being through bio-accumulation processes ( for detecting/measuring POPs in various compartments of environment for exposure monitoring and to POPs risk evaluating on human and wild-life at regional and global scales. Passive sampling gained considerable attention for their continuous sampling of contaminant from environment and also catching episodic exposure in water (Schäfer et al. 2008, Shaw &Mueller 2009). However, studies along the Indus river Pakistan are still scarce, and no study has done so for on passive sampling of water to measure the dissolved POPs fraction in the surface water. Similarly, the authenticated data from Pakistan on dissolved fraction of POPs is also needed as previously studies followed questionable method of water sampling and/or they collected the water samples including the suspended particulate material, which overestimate the concentrations of "dissolved" POPs. The problem can be overcome by following few recent studies from USA and European labs, which have described the use of these novel passive samplings (i.e., LPDEs and PS, PUFs) for detecting/measuring POPs in different water bodies (Harner et al. 2006, Levy et al. 2009, Lohmann &Muir 2010. Additionally, open burning of e-waste has also been reported to release a huge amount of PCBs and other ame retardants (Novel brominated ame retardants (NBFRs) and organophosphorus ame retardants (OPFRs) ( that there is lack of the reliable data, different inventories and research studies addressing e-waste issues in Pakistan and which is one of the major source of PCBs in different areas of country. The present study was conducted for reporting the levels, distribution and apportionment of emission sources of POPs (i.e., organochlorine pesticides (OCPs), polychlorinated biphenyls (PCBs) into the Indus River, Pakistan. The study has aimed on the following main objectives. 1) To investigate the levels, compositional pattern, potential sources, and risk assessment of POPs into the Indus water ecosystem of Pakistan via passive sampling. 2) To study the air-water exchange and cycling of POPs into ambient air and surface water of Indus River.

Study Area and Field Strategies
The Indus ood-plain area is inhibited by more than 100 million human people and these areas is dependent on Indus River for agricultural activities and for shing. The Indus River originates in the Tibetan plateau near the Lake Mansarovar, take a course via the Ladakh region (Jammu and Kashmir), Gilgit Baltistan and then ows towards south direction along entire length of Pakistan and then merges into Arabian Sea near Karachi, Sindh Pakistan. Indus River is the largest as well as longest river that ows throughout the entire length from north to south of Pakistan and its catchment area possesses diverse ecological systems i.e., snow capped northern areas, lower Himalyian mountainous valleys, agricultural plains in the south and coastal areas near Arabian sea. The Indus River also runs through the densely polluted and populated industrial cities of Punjab Province and the Khyber Pakhtunkhwa (KPK) Province, Pakistan, receiving a greater amount of pollutants from the industrial and municipal e uents through various streams and its tributaries such as Chenab River, Ravi River and Jhelum River.
After the preliminary survey of study area, fteen sampling sites were selected along the entire stretch of Indus River, where the passive samples were installed during 2014. These sites include Gilgit, Hunza, Skurdu, Swat (Gullibagh, Kalam, Malamjaba), Swabi, Nowshera Mianwali, Bakkar, Layyah, D.G Khan, Khairpur and Hyderabad. The detailed description of each site is given as Text S1 and Table. S1. Sites location was identi ed in the catchment using a GPS (Global Positioning System, Garmin) and mobile weather station was used for recording the climatic data of each studied site. Low Density Polyethylene sheets (LDPE) were deployed in the surface waters for three weeks. LDPE passive water samplers were deployed 1 m below the surface attached to small buoys. Each sampler was assembled at the deployment sites for avoiding contamination in transportation. At the completion of the deployment period, the LPDE were retrieved, resealed and returned to Lohmann Lab, University of Rhode Island, and stored frozen until further analysis.
Passive sampling for water First, a roll of about 2 cm wide, 60 cm long and 100 microns thick polyethylene (PE) tubing was cut off and cleaned with hexane. They were treated twice for 24 h in graduated cylinder with 250 mL n-hexane and were cleaned from inside and dried in a nitrogen stream for 30 minutes. Subsequently, these tubes were welded (front and back twice) and put for storage in container. For the spiking, two performance reference compounds (PRCs) such as TCmX and PCB-29 were selected which are deuterated possible structural isomers of the substances. For this step, a polyethylene tube was put in 1.5 L bottle containing 1 L of water and then added 10 µg of deuterated substance.
These were rst capped and clamped for 72 hours in over headed shaker (15 rpm speed). Discarded the solution and the tubes dried with the pulp and frozen till outdoor exposure. For the deployment, the polyethylene tubes were then placed in space provided by the grating and screwed. The grill was attached with the steel hook in water and give exposure for up to three weeks. A photo was taken for documentation, quality assurance (Fig. S4). The eld blanks were carried for capturing the background contamination. The grills were removed after three weeks. The detailed methods have been reported elsewhere (Schäfer et al. 2010).
Extraction and Clean-up of passive water samples Details of LDPE extraction can be found elsewhere (Khairy &Lohmann 2013). Brie y, the PE tubes were cold extracted twice in ethyl acetate each for 24 hours after spiking with 10 µL of surrogate standards PCB-209 and TCmX for calculating method recovery. The resulting extracts were combined and concentrated to 1 mL on a turbovap, solvent exchanged to the hexane, and concentrated to ~ 50 µL using a gentle stream of ultrapure nitrogen. Injection standard was added for volume correction to all sample extracts prior to analysis. . The injector temperatures were set to 250°C. The initial temperature of oven were set at 150°C for 3 min, and then raised to 290°C at a rate of 4°C/min and held for 10 min. Persistent organic pollutants (POPs) were then determined in selected ion mode (SIM). MSD sources and quadrupole temprature were 230°C and 150°C. The MS used in this case were in SIM mode having two ions monitored for each of the target compound group in a speci c window. DDT inlet degradation was daily checked and control within the 15%. All the persistent organic pollutants (POPs) were quanti ed using HP-Chemstation to con rm the peaks.

Statistical Analysis
Analytical results were organized to form a multielemental database by using Excel software. Graphical representation of POPs data was made using Microsoft Excel 2016 and Arc-GIS (version 10.2) was used for showing various sampling locations on map and the spatial distribution of different chemicals. A level of p < 0.05 was considered statistically signi cant for multiple comparisons. A one-way ANOVA then followed by Tukey's HSD post hoc test were employed for making multiple comparisons among the different contaminants and zones. Principal component analysis/Multilinear Regression (PCA/MLR) by using SPSS Statistics (version 21) was then performed for investigating the qualitative source for target OCPs and PCBs measured in the water of Indus River, Pakistan. The detail of calculations of Air-water exchange gradient of OCPs and PCBs, water sampling rate, and human health risk assessment is describes as Text S-2, S-3 and S-4 in the supplementary material.
Quality assurance & Quality control (QA/QC) All the samples were rst subjected to have strict quality control measure. For QA/QC, all the instruments were daily calibrated with the calibration standards. Solvent blanks, the eld blanks and the procedural blanks were analyzed by using the same procedure as for the real samples. For every batch, internal standards were then injected into Gas Chromatogram before the sample injection. Moreover, surrogate standards (PCB-209 and TCmX) were also added to all the samples (LPDE) before installation in the eld for monitoring matrix effects and procedural performance. Mean surrogate recoveries calculated for the PCB-209 and TCmX, ranging between 70-105% and 60-75% for PCB-209 and TCmX. Reported concentration of OCPs and PCBs were corrected to surrogate recoveries. DiBB (2,5 dibromobiphenyl), TeBB (2,2'5,5' tetrabromobiphe), PeBB (2,2',4,5',6 pentabromobi) and OCN added before GC-MS analysis as an internal standard. The processing of data acquisition was controlled by Agilent MSD Productivity Chemstation software. All the chemicals were of analytical grade and were purchased from the Merck, Germany. All solvents used in this study, redistilled for purifying solvents to reduce any interference effects of solvents. All glassware used in this study were baked (450 ºC) for 6 hours prior to use.

Freely dissolved OCPs and PCBs
Sampling rates (Rs) for water passive samplers are given in (Table S2). The value of sampling rate for polyethylene passive water samplers were ranging from 7. Freely dissolved OCPs levels ranged from 34-1600 pg/L throughout Indus River as shown in Table 1 & Fig. 1. Among the studied OCPs, DDTs (0.48-220) were dominated, followed by endosulfan (0.20-14), HCBs (0.23-6.4), chlordane (0.09-4.58) and HCH (BDL-2.64). p,p′-DDE accounted for higher frequency among the studied OCPs at most of studied site locations. In contrast, trans nonachlor (0.09 pg/L) was among the less frequently detected pesticide into the surface water of the Indus River. The freely dissolved levels of OCPs in various zones showed the trends as follows; Alluvial Riverine zone (ARZ)   (Gao et al. 2013, Jiang et al. 2009). In present study, p,p′-DDT /( p,p′-DDE + p,p′-DDD) value is lower than unity at most of the studied location with few exceptions included the sites of Swat, Hunza and Gilgit of FMZ, where DDTs diagnostic ratios indicating the current as well historical usage of this banned pesticide. Additionally, the values for the DDD/DDE ranging from 0.09-5.95 (with mean value of 0.65). Generally, all the sites showed the values below unity and re ected the aerobically degradation of DDTs in these areas, might be due to the frequent ploughing/crop cultivation in the agricultural eld. In contrary, Hunza site showed the DDD/DDE as 5.95 and indicating that the DDT in the soil is metabolized in anaerobic environment, which can be explained by high organic matter contents and its microbial degradation in these forests covered areas. The o,p′-DDT/p,p′-DDT ratio has reported to assess the usage of dicofol as possible additional source of DDTs in the environment. The ratio of the o,p′-DDT / p,p′-DDT ranged from 0.2 to 0.3 in the technical DDT and is 7.0 ± 2.2 in dicofol (Qiu et al. 2005). In the current study, ratio of the o,p′-DDT / p,p′-DDT ranged from 0.23 to 0.98 and highlighted that the major usage of technical DDT in the Indus ood plain (Fig. S2).
Among the HCHs, γ-HCH contributed signi cantly among all dissolved HCHs and ranged between 0.20-2.34 pg/L. These trends for HCHs occurrence into the surface water of Indus water can be justi ed due to higher water solubility of γ-HCH. Moreover, lindane (containing 90% of γ-HCH) has also been reported to be used in current study areas (Eqani et Table 2. The occurrence trends of different PCBs homologues into Indus River follows the order as: tri-CBs (~ 29-71%) > penta-CB (~ 3-9%) > tetra-CB (~ 3-8%) > octa-CBs (~ 3-6%) > hexa-CBs (~ 2-3%) > hepta-CBs (~ 1-2%) (Fig. 2b). The present study has shown that levels of PCBs at FMZ, WMZ and ARZ were consistent with the levels reported for Lake Erie (52- (Table S5). Among the individual mono-ortho DL-PCBs and the non-ortho DL-PCBs, PCB-105 and − 169 contributed very signi cantly toward total calculated TEQs. Higher TEQs for mono-ortho DL-PCBs were calculated at ARZ followed by LLZ, FMZ and WMZ which showed the higher industrial activity at ARZ. Principal component analysis/multiple linear regression (PCA/MLR) describes and identi es between the related and the unrelated source tracer and also percentage contribution from the various sites along the Indus River, Pakistan. PCR/MLR was performed on the data of OCPs and PCBs for individual metabolites of DDTs, HCHs and PCBs homologues (i.e., tri-CB, tetra-CB, penta-CB, hexa-CB, hepta-CB and octa-CB) using varimax rotation method (Fig. S7).
In frozen mountain zone (FMZ), the PCA extracted four components (explaining 33.33, 26.18, 23.80 and 13.02 variability, respectively) and also shown as Table S8. The PC-1 accounted for 33.33 % variance and highlighted the p,p′-DDD, o,p′-DDT, o,p′-DDE, and p,p′-DDE. In the past, DDTs were used for agricultural purposes and malarial control program and PCA-MLR has also indicated the historical as well as current illegally use of DDTs in agriculture and health sector in the studied area. Diagnostic ratios for DDTs in these areas has also indicated its fresh and/or aged uses of DDTs in these areas. Moreover, wet and dry deposition of atmospheric dust particles through the process of altitudinal fractionation can also justify the DDTs contamination in these areas (Khan et al. 2017, Sohail et al. 2018). Atmospheric DDT fallout into the freshwater ecosystem has also been reported worldwide as > 150 Kg ΣDDT dust borne contamination has been estimated to the Great Lakes, USA and which was due to wet and dry deposition (Hoff et al. 1996). The PC-1 also dominated by various polychlorinated biphenyles including octa-CB, hepta-CB and penta-CB, which re ected the long range dust borne PCBs contamination into this area For the low-lying zone (LLZ), we considered PC-1 which explained more than 90% of the variance in data obtained from the Indus River and highlighted a mixed source containing of pesticides and electronics. Agricultural application of pesticides as well as industrial activities are the main reasons for the contamination of this zone. All districts included in LLZ zones are characterized by harsh climatic conditions (i.e. wind storms, high temperature, less vegetation cover and low rainfall) and historical agricultural background. Moreover, these low-lying areas are also receiving industrial, agricultural and urban waste from other parts of the country via local stream network and Indus river system, which may further ush out into Arabian sea or re-emitted through volatilization ).

Air-Water exchange gradients of OCPs and PCBs
In general, considering zone wise results of air-water exchange of OCPs, there were net deposition of DDT isomers (p,p / -DDT) and its metabolites at all studied zones, which might be due to its current illegal use in the region for crop protection and malarial control, and resulted into elevated levels of atmospheric OCPs and nally deposition into surrounding water bodies. However, few exceptions of net volatilization of DDTs metabolites into FMZ highlighted its re-emission from POPs repositories (via historical cold trapping) at various colder sites as also described by (Chen et al. 2008 (Table. S4), suggesting low ecological risk associated with the levels of OCPs/PCBs into the Indus River. However, it is well known that these chemicals have potential to bioaccumulate due to their lipophilic characteristics (log K ow ) and even at very low levels into the fresh water system, contributed signi cantly towards the total body burdens into the aquatic organisms ( sh) and subsequent food chain and nally associated humans.
We calculated the chemical daily intake (CDI) estimates for the assessment of intake of chemicals via various routes (i.e., ingestion and dermal) contact into the associated human body. adding cumulative risk of OCPs/PCBs into the associated exposed population (USEPA, 2001). Furthermore, exposure to low-level of OCPs/PCBs (oral or dermal) by the drinking water may also result into carcinogenic effects into human which is an indicator for potential cancer risk (CR). If CR is greater than 10 − 6 then CR exists in the exposed population due to OCPs/PCBs-contaminated water intake (US- Although, it has been observed that their no/low known carcinogenic risk from both OCPs and PCBs through oral and dermal exposure of water from the different studied zone in Indus River (Fig. S5 &   S6). Nevertheless, according to recent report describing the effects of toxic stimuli combinations on determination of safe exposure limits has been published and concluded that there is no reason to believe today that any of the Exposure Limits on potentially toxic stimuli that have been set by any of the regulatory agencies are fully protective against serious adverse health effects (Kostoff 2018). Moreover, this study highlighted the contribution of water matrix via non-carcinogenic and carcinogenic risk, which in addition to dust, soil and biota (diet), may pose several health problems to associated human population.

Conclusion
This is the rst study reported authenticated data of dissolved OCPs and PCBs measured through the passive water sampling by using polyethylene sheets (LPDE) throughout the Indus River, Pakistan. In general, the Indus River water is being moderately contaminated with OCPs and PCBs when compared with the other water bodies throughout the world. Our zone-wise results of air-water exchange of OCPs and PCBs re ected the cycling of POPs (volatilization and deposition) along the Indus River, Pakistan, which may be due to rapid land use, abruptive climate change and regional atmospheric in uences. Our results also highlighted the LRAT of POPs into northern colder areas Pakistan and highlighted the reemission of POPs from their repositories (via historical cold trapping) at various colder sites. These POPs likely entered the Indus via melting of glaciers and ice masses and which also continually volatilizing the POPs into ambient air of these pristine areas. Nevertheless, local sources also contributed towards the recent emission of these toxic chemicals (especially DDTs), which are being used for agricultural and industrial purposes and transported/deposited into the entire geographic zones along Indus river ecosystem by wet/dry deposition process.   Compositional pro le of freely dissolved (a) DDTs and (b) PCBs along the Indus River, Pakistan. Note:

Declarations
The designations employed and the presentation of the material on this map do not imply the expression of any opinion whatsoever on the part of Research Square concerning the legal status of any country, territory, city or area or of its authorities, or concerning the delimitation of its frontiers or boundaries. This map has been provided by the authors. Air-Water exchange of DDTs (a) and PCBs (b) along Indus River Flood-plain, Pakistan.

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download.