Occurrence of organic micropollutants and human health risk assessment based on consumption of Amaranthus viridis, Kinshasa in the Democratic Republic of the Congo

The contamination of water resource and food chain by persistent organic pollutants (POPs) constitutes a major environmental and human health concern worldwide. The aim of this study was to investigate the levels of POPs in irrigation water, soil and in Amaranthus viridis (A. viridis) from different gardening sites in Kinshasa to evaluate the potential environmental and human health risks. A survey study for the use of pesticides and fertilizers was carried out with 740 market gardeners. The levels of POPs (including organochlorine pesticides (OCPs), polychlorinated biphenyls (PCBs), polybrominated diphenyl ethers (PBDEs), and polycyclic aromatic hydrocarbons (PAHs)) were analyzed in irrigation water and 144 vegetable samples collected from different gardening sites. The assessment of potential human health risk was estimated by calculating daily intake and toxic equivalency to quantify the carcinogenicity. The results show highest PAH levels in A. viridis from all studied sites. The concentrations of the sum of seven PCBs (Σ7PCBS) congeners in analyzed plants ranged between 15.89 and 401.36 ng g−1. The distributions of OCPs in both water and A. viridis were congener specific, chlorpyrifos-ethyl and p,p′-DDE were predominantly detected. Among PBDEs, only BDE47 was quantified with noticeable concentration in A. viridis, while no PBDEs were detected in irrigation water. Higher estimated daily intake values indicate that consuming leafy vegetables might associate with increased human health risks. However, calculated incremental lifetime cancer risk values indicates no potential carcinogenic risk for the local population. The results of this study provide important information on A. viridis contamination by POPs and strongly recommend implementing the appropriate measures to control the use of chemicals used in studied gardening areas. Thus in Kinshasa, urban agriculture control programs for POPs and fertilizers is very important in order to protect the public health through direct and dietary exposure pathways.


H I G H L I G H T S
• Survey study was performed for the use of pesticides and fertilizers. • Persistence Organic Pollutants (POPs) in soil, water and A. viridis were investigated. • Highest PAH levels were observed in soil, water and A. viridis from all studied sites. • Estimated daily intake indicates potential health risks associated to leafy vegetables consumption. • Incremental lifetime cancer risk values indicate no potential carcinogenic.

G R A P H I C A L A B S T R A C T a b s t r a c t a r t i c l e i n f o
The contamination of water resource and food chain by persistent organic pollutants (POPs) constitutes a major environmental and human health concern worldwide. The aim of this study was to investigate the levels of POPs in irrigation water, soil and in Amaranthus viridis (A. viridis) from different gardening sites in Kinshasa to evaluate the potential environmental and human health risks. A survey study for the use of pesticides and fertilizers was carried out with 740 market gardeners. The levels of POPs (including organochlorine pesticides (OCPs), polychlorinated biphenyls (PCBs), polybrominated diphenyl ethers (PBDEs), and polycyclic aromatic hydrocarbons (PAHs)) were analyzed in irrigation water and 144 vegetable samples collected from different gardening sites. The assessment of potential human health risk was estimated by calculating daily intake and toxic equivalency to quantify the carcinogenicity. The results show highest PAH levels in A. viridis from all studied sites. The concentrations of the sum of seven PCBs (Σ7PCB S ) congeners in analyzed plants ranged between 15.89 and 401.36 ng g −1 . The distributions of OCPs in both water and A. viridis were congener specific, chlorpyrifos-ethyl Science of the Total Environment 754 (2021) 142175
Three major types of POPs are commonly reported in the environment for many years. They were mostly anhropogenically derived compounds including polychlorinated biphenyls (PCBs), polyaromatic hydrocarbons (PAHs), and organochlorine pesticides (OCPs). Moreover, according to the Stockholm Convention 2004 regulations, protection of environment and human health risk from POPs is a high priority (Stockholm Convention, 2004). For instance, through a variety of food matrices (e.g., vegetables, eggs, fish, meat, oils, and milk) consumption, these contaminants were reported to induce health effects such as neurotoxicity, endocrine disruption, cancer, reproductive disorders, leukemia, asthma and health risks to fetal development (Lü et al., 2014;UNEP/WHO, 2013;Gilden et al., 2010;Kima et al., 2017;Fernandes et al., 2019). POPs are very persistence in soil and can affect crop quality and yield. Consequently, many studies propose the organic pollutant degradation mechanism pathways (such as photocatalytic degradation) and the remediation efficiency of multi-element contaminated soil in order to reduce exposure, guarantee food safety and protection of human health (e.g. Venny et al., 2012;Ye et al., 2020Ye et al., , 2019aYe et al., , 2019bWeber et al., 2019).
Vegetable is an essential part of the human healthy diet and considered as a source of many essential nutrients to maintain normal physiological functions, antioxidants, dietary fiber metabolites and to prevent several diseases (Boeing et al., 2012;FAO/WHO, 2018;Azi et al., 2018). Vegetables also attract a wide range of pests and affected by diseases, and therefore need intensive pest management (Dinham, 2003). Leafy Amaranthus spp. are very important to the human diet because its constitute excellent sources of magnesium, manganese, phosphorus and vitamin C, its contain higher mineral levels than many common leafy vegetables, have a calorific value of 43.35 kcal, crude protein 2.11%, moisture content 87.90%, carbohydrate content 7.67%, crude fiber 1.93%, crude fat 0.47%, and ash content 1.85% (Sharma et al., 2012;Mota et al., 2016;Jiménez-Aguilar and Grusak, 2017). Amaranthus spp. are among of the most domestic vegetables consumed in south of Asian and sub-Saharan African countries, and currently imported and sold in African and Asian shops located in many EU countries. The local population consumes daily and during the big festivals A. viridis leaf generally cooked and favors it for taste and tradition. The plant is mainly cultivating in peri-urban market gardening (Akinola and Eresama, 2009;Islam and Hoque, 2014;Azi et al., 2018;Bashri et al., 2016).
In the DRC, particularly in Kinshasa its Capital City, the urban agriculture (market gardeners) plays an economic and social role in daily life of the population and provides more than 60% of the consumed fresh produce supply of the city. After cassava leaf, A. viridis has been identified to be a second most consumed leaf vegetable in the DRC. The A. viridis cultivation is mainly performed in the peri-urban municipalities near riverbanks. This activity supports many families in Kinshasa and employs more than ten thousands of people. It can therefore contribute to the sustainable development of the city under certain conditions, especially through its professionalization, the non-use of chemical inputs and the equitable distribution of arable land (Musibono et al., 2011). Since the year 2013, the Ministry of Agriculture of DRC has approved the import, sale, and use of several pesticides including endosulfan, dithiocarbamate, rayasansulfan, rhodiatox, delthaméthrine and metyldor for urban agricultures (M.A.P.E., 2016). Consequently, pesticides and fertilizers are extensively applied for pests' control to improve yield of urban agriculture practices for the production of vegetables and fruits. On other hand, there is no control or application of regulation concerning the use of these substances. Pesticides use and atmospheric deposition have been considered as the major sources of POPs in plant products (Polder et al., 2010). Besides, the species and physiology of the plant can also affect the accumulation of POPs from environment matrix (Sun et al., 2019). Thus, vegetables contaminated with POPs present a potential health risk for the consumers. Nevertheless, there is lack of information on POPs accumulation in plants particularly leafy vegetables in Democratic Republic of the Congo (DRC). A very few studies reported the use of pesticides and accumulation of OCPs, PCBs and PBDEs in vegetables (such as tomatoes and cabbages) collected from urban areas of the provinces of Kongo Central and Bukavu in DRC (Muliele et al., 2017;Kavatsurwa et al., 2014).
To our best knowledge, the studies on the accumulation of POPs in this local consumed and exported vegetable (A. viridis cultivated in Kinshasa is exported to several African and EU countries), which attract high applications of authorized and non-authorized pesticides as well as fertilizers are still scarce. Therefore, the aim of the research presented in this paper was to investigate the levels of organochlorine pesticides (OCPs), polychlorinated biphenyls (PCBs), polybrominated diphenyl ethers (PBDEs), and polycyclic aromatic hydrocarbons (PAHs) in irrigation water, soil from gardening areas, and in A. viridis in order to evaluate the potential environmental and human health risks. This research was performed in July/August 2018, in eight main gardening sites in Kinshasa, the capital city of the Democratic Republic of the Congo. The levels of POPs in A. viridis were coupled with daily intake data in order to assess the local consumer health risks.

Survey study of pesticide use
The survey study for the use of pesticides and fertilizers was performed in 8 studied sites as a questionnaire. A total of 740 gardeners (100/site except for the Kimpoko site (40)) were interviewed concerning the types and practices of pesticides and fertilizers use (pesticides and fertilizers use practices, application interval and water method used, pre-harvest packaging and feeling of discomfort related to the application of pesticides). All interviewed gardeners are working in sampled sites; 54.6% males and 45.4% females, with an average age of 47 years.

Study site description
This research was performed in 8 gardening sites located in Kinshasa, the capital city of DRC (Fig. 1). The sampling sites are named and labelled as; Rifflaert (RI), Lemba-Imbu (LI) and Cecomaf (CE) nearby N'djili River, Kimpoko (KI) nearby Kimpoko River, Tshuenge (TS) nearby the Tshuenge River, Agricole-Mombele (MO) nearby Limete River, Monastery (MON) nearby Funa River, and Saïo (SA) nearby Saïo avenue. These sites were selected according to the high surface and cultivation frequency of A. viridis. The GPS coordinates and activity performing nearby sampling sites are reported in Table 1S (Supplementary Data (SD)).

Water, soil and plant sampling
The water, soil and plant sampling took place in July/August 2018. Water samples (500 mL sealed in clean plastic bottles) were manually collected in triplicate from each river (near the crop fields) used for irrigation, except for Saïo site (where water samples were collected from a well used for irrigation). Soil and plant sampling was performed in the crop fields, when the vegetable reached the stage of harvest ( Fig. 1S-A, in SD). For good representability, soil and the vegetable samples from 6 sub-sites (or sampling points) within the main selected agricultural fields (8 sites) were collected using classic four quadrat sampling approach (Bhatia et al., 2015;Islam and Hoque, 2014;Chen et al., 2010). From each sampling point, the collected sub-samples (about 500 g of edible plant parts (leaf)) were thoroughly mixed to form composite samples that were investigated in this study. About 150 g of rhizospheric soil (0-30 cm) were collected in triplicate (simultaneously with vegetable samples). For their conservation after sampling, water and soil samples were stored at 4°C and immediately transported to the laboratory for pre-treatment and analysis within 72 h. Before analysis, edible plant parts (leaf) were washed with deionized water, weighted, lyophilized and water content was calculated. Soil samples were sieved through a 1 mm mesh size sieve, weighted, lyophilized and water content was calculated.

Chlorinated pesticides, PCBs, PAHs and PBDEs analysis
Chlorinated pesticides, PCBs, PAHs and PBDEs analysis in water, soil and plant samples was performed using gas chromatography with triple mass spectrometry detection (GC-MS/MS and UPLC-MS/MS, Thermo Scientific, TSQ Quantum XLS Ultra, Waltham, MA, USA) as performed in our previous studies (Lehmann et al., 2017(Lehmann et al., , 2018Kilunga et al., 2017;Mwanamoki et al., 2014;Thevenon et al., 2013) (details in Supplementary Data). The QuEChERS extraction procedure, quality control and quality assurance were performed as described in our previous studies (Lehmann et al., 2018(Lehmann et al., , 2017Kilunga et al., 2017;Mwanamoki Thevenon et al., 2013 (SD)). The validation of the used analytical methods was performed by calculating the limit of detection (LOD), the limit of quantification (LOQ), and recovery values as performed by Lehmann et al. (2018). The limits of quantification (LOQ) for the selected target analytes, material and chemical sources, purity as well as operating parameters for GC-MS and UPLC-MS/MS are described in Supplementary Data (sections S1 and S2, Tables S2-S5).
2.5. Assessment of potential health risks 2.5.1. Daily intake of PAH The daily intake of PAH (DI-PAH) through consumption of PAHcontaminated vegetable was determined using the following equation: where CPAH, C factor , V intake and B weight represent the PAH concentration in vegetables (μg kg −1 ), conversion factor, daily intake of vegetables, and average body weight, respectively. To convert fresh weight of vegetable into dry weight, a conversion factor (0.085) was used (Rattan et al., 2005). Average daily vegetable intake for adults were considered as 0.067 kg person −1 day −1 (PNUD-SOCOGEM, 2000; Bonkena et al., 2018), while the average adult body weight was considered as 57.8 kg (Mbenza et al., 2006).

Dietary exposure and cancer risk of PAHs
Toxic equivalency factors (TEFs) were used to quantify the carcinogenicity of selected PAHs relative to BaP (e.g. Kang et al., 2017;Ali, 2019;Gong et al., 2019). The calculated TEFs were taken as 0.001 for Nap, Flo, Phe, Flu, Ace and Pyr; 0.01 for Ant, Chr, BghiP; 0.1 for BaA, BkF, BbF, and 1.00 for BaP (Mehdinia et al., 2015;Mocek and Ciemniak, 2016;Gong et al., 2019;USEPA, 2002). Total toxic BaP equivalent (TEQ) for PAHs was calculated using the following equation: where TEF i is the corresponding toxic equivalency factor for PAHs and C i is the concentration of the individual PAH.
Incremental lifetime cancer risk (ILCR) of PAHs from vegetable consumption was calculated using the following equation: where ILCR refers to the incremental lifetime cancer risk of the dietary exposure, IR is the daily PAH exposure level (g day −1 ), EF (365 days year −1 ) is the exposure frequency, BW = the average body weight of population (57.8 kg), SF is the oral cancer slope factor of ben (a)pyrene (7.3 mg kg −1 day −1 ), CF is a conversion factor (10 −-6 mg ng −1 ), AT = average time (equal to 75 years for carcinogens) and ED is the exposure duration (70 years) (USEPA, 1993(USEPA, , 2002.

Statistical analysis
The statistical treatment of the data has been realized using SigmaStat 11.0 (Systat Software, Inc., USA). The data were subjected to the Spearman's Rank Correlation test to investigate the possible positive and negative relationships among variables. Principal Component Analysis (PCA), a multivariate statistical analysis was performed using R (R Core Team, 2015) in order to understand relationship among analyzed compound and their potential sources. Prior to performing PCA analysis, data were centered in order to maximize the dispersion (Kilunga et al., 2017).

Surveys on the use of pesticides and fertilizers
We have surveyed the use of commonly used pesticides and fertilizers in the study sites. Based on the survey, the use of pesticides is as follows: Thiodan endosulfan sulfate (94.6%) > dithiocarbamate (37.2%) > rhodiatox (4.7%) > cypermethrine (4.5%) > coga 80 WP-mancozebe (2.7%) > ivorymancozebe (2%) > karate-lambda cyhalothrine (1.5%) > pacha-lambda cyhalothrine (1.2%) > delthamethrine (0.8%) > and DDT (0.8%). The higher use of thiodan endosulfan sulfate can be explained by its low cost, frequent availability in the Congolese market and the lack of regulation of this pesticide. This result is consistent with results recently obtained by Ngweme et al. (2019), who found that endo sulfan sulfate was the most commonly used pesticide in Kinshasa. There are only nonprofessional farmers. About 25.4% of interviewed farmers don't know what they are exactly used as pesticides and 90% don't know what they are using as fertilizers. This leads improper and large use of pesticides and fertilizers released in the environment. The quantity and pesticides practices (by spraying) depend on the estimation of each farmer ( Fig. 1S-B, in SD). Market gardeners refer to the advice of other farmers or suppliers and use is also influenced by climatic and socio-economic status (Ngweme et al., 2019;Olisha et al., 2020).

Distribution of persistent organic pollutants (POPs) in irrigation water
Results of PCB, OCP, PBDE and PAH concentrations in analyzed irrigation water samples collected from different locations at Cecomaf (CE), Rifflaert (RI), Lemba-Imbu (LI), Tshuenge (TS), Kimpoko(KI), Monastery (MON), Agricole-Mombele (MO) and Saïo (SA) gardening sites are presented in Tables 1a and 1b. Noticeably, concentrations of analyzed PCBs, PBDEs and OCPs (except for chlorpyrifos-ethyl in the site KI and MO) in irrigation water were always below LOQ in all samples. In irrigation water, PAHs were observed to be the most predominant chemicals detected. The concentration of PAHs (ng L −1 ) were ranged from 31.9 to 452.9; 9.6 to 879.9; 9.2 to 675.7; 4.4 to 326.7; 8.5 to 1310.4; 8.9 to 344.2; 37.0 to 242.5 and 8.5 to 183.2 for KI, MO, SA, MON, RI, CE, TS and LI sampling sites, respectively. The spatial variation in PAHs concentrations in water samples suggests that the contamination of PAHs in the study sites could be attributed to urban activities. The highest Σ16 PAHs was noted at the site RI with the value of 2673.7 ng L −1 while the lowest was recorded at the site LI with the value of 395.8 ng L −1 . The results indicate that over 80% of the water samples in each study sites are contaminated by the target PAHs. Water samples collected at the site KI recorded high levels of carcinogenic PAHs. The increased level of PAHs in water samples from KI site was probably due to grass, wood, or coal combustion and petrogenic activities. Among analyzed PAHs, low molecular weight (LM) chemicals such as naphthalene and phenanthrene were predominantly detected in water samples could probably be linked to the urban land use. Regarding the OCPs targeted in this study, only chlorpyrifos-ethyl, p,p′-DDE, and p,p′-DDD were detected in some sites. p,p′-DDE, was found to be the most dominantly occurring pesticide in irrigation water. The levels of analyzed POPs in irrigation water from our study sites were generally higher in comparison with the data obtained in a control site (Lake Ma Vallée), where the values of POPs and toxic metals in the water column and sediments are in many cases below detection limit (Mwanamoki et al., 2014;Laffite et al., 2020).

Level of persistent organic pollutants (POPs) in A. viridis
The sum of twelve congener PCBs (Σ 12PCBs) in A. viridis is presented in Table 2a. The concentrations of PCBs were reported in ng g −1 dry weight. The occurrence similar to those of irrigation water was also observed for POPs in A. viridis. The concentrations of the 12 PCBs congener were ranged from 0.67 to 114.76; 0.35 to 93.79; 0.20 to 95.14; 0.68 to 71.76; 0.05 to 4.23; 0.05 to 7.88; 0.36 to 97.19 and 0.06 to 2.78 for KI, MO, SA, MON, RI, CE, TS and LI sampling sites, respectively. In general, PCB congeners in A. viridis were dominated by 101, 118, 138 and 153 and varied significantly among the sampling sites. The highest concentration of ΣPCBs (401.36 ng g −1 dw) was found at the site KI and the lowest levels were measured at the site LI (15.49 ng g −1 ). The high levels of ΣPCBs in A. viridis were probably due to uncontrolled waste disposal and incineration, industries emissions, atmospheric deposition and land use in the studied area. The PCBs levels were much higher than those previously reported in leafy vegetables from South Africa (Olatunji, 2019).
The concentrations of PAHs in A. viridis are presented in Table 2a. Of the targeted PAHs, fifteen PAHs congeners were detected in all A. viridis samples. The temporal variation of total PAHs (Σ16 PAHs) concentrations ranged between 93.7 and 469.1 ng g −1 dw for all of the vegetable samples analyzed. Significant differences in PAHs concentrations were observed in A. viridis corresponding to sampling locations. Acenaphthylene was not detected in all plant samples. The concentrations PAH congener Fluoranthene (140.4 ng g −1 dw) in vegetable was found to be the highest while the concentrations of PAH congener dibenz(a,h)anthracene was found to be the lowest (0.1 ng g −1 dw).
It was reported by Yunker et al. (2002) that PAHs can originate from natural or anthropogenic processes. Four diagnostic ratios were calculated in this study to better understand the potential sources of PAHs including wood and coal combustion, traffic emissions, petrogenic, and pyrolytic. Using the ratio between the concentration of Fluo/(Fluo + Pyr) or IDP (IDP + BghiP), it is possible to determine if PAHs are from petrogenic or pyrogenic origin. In the first case, if the ratio is <0.4, the source is petrogenic, in the second case, when it is between 0.4 and 0.5 the source is associated with petroleum combustion, and when it is >0.5, the source is grass, wood, or coal combustion (Yunker et al., 2002;Manneh et al., 2016). According to Manneh et al. (2016), the ratio of IDP/(IDP + BghiP) < 0.2 represents the source is petrogenic; a ratio ranged from 0.2 to 0.5 is considered as the source of petroleum combustion; a ratio greater than 0.5 indicates grass, wood, or coal combustion. In addition, the ratio of BaA/(BaA + Chry) smaller than 0.2 or 0.35 indicates either a petroleum or combustion source and a ratio greater than 0.35 indicates pyrolytic origin (Manneh et al., 2016). Furthermore, according to Budzinski et al. (1997), if a ratio of Phen/ Anth > 10 the source is petrogenic, otherwise, if the Phen/Anth < 10 the source is pyrolytic. According to the above ratio calculated for irrigation water (Table S6 in SD) from KI and RI, PAHs probably have originated from petrogenic processes and/or pyrolytic sources, while PAHs found in SA, TS and LI irrigation water have originated from pyrolytic sources. On the other hand, PAHs found in MO could originate from grass, wood, coal combustion or pyrolytic sources. Finally, for the site CE, PAHs could have originated from grass, wood combustion or petroleum source.
Regarding the A. viridis sample, the calculated ratio indicated above shown for sites KI, MO, SA, TS and LI, PAHs identified could originated from grass, wood, petroleum combustion or pyrolytic processes, while for the sites MON, RI and CE, PAHs have originated from grass, wood, petroleum combustion or petrogenic processes. It should be noticed that people who are living around the studied sites burn shrubs and habitants use coal and wood to cook. This combustion of coal and wood could explain the origin of some PAHs around the sites. Besides, some of these sites, such as Cecomaf and Saïo, are located along the main roads with heavy traffic, which releases fuel combustion products.
The results of the OCP concentrations are shown in Table 2b. Similar to pesticides in irrigation water A. viridis contain Chlorpyrifos-ethyl, p,p ′-DDE, o,p′-DDE, p,p′-DDD and p,p′-DD. A similar concentration of DDTs was recorded in this study compared to those of previous studies in African countries (Ndengerio-Ndossi and Cram, 2005;Kolani et al., 2016). High levels of chlorpyrifos-ethyl (23.05 ng g −1 dw) found in A. viridis collected from the site TS implying intensive use of pesticides in agriculture practices. Moreover, the Ministry of Agriculture of DRC has approved the import, sale and use of chlorpyrifos for insecticidal applications in agriculture since the year of 2013 (M.A.P.E., 2016). Among the DTTs, p,p′-DDE, and p,p′-DDT was found to be the most frequent chemicals in most of the samples collected from the study areas. The most highly contaminated p,p′-DDE plant sample was collected from the site CF. The results of DDT pesticides in plant samples might be influenced by the level of water and soil contamination. On the other hand, the high levels recorded for DDTs, which was already banned in agriculture use in many African countries, was still probably in illegal use on agriculture crops (Kilunga et al., 2017). The frequent use of DDT in malaria vector control in the studied regions could also attribute the higher pesticides occurrence in A. viridis. Remarkably, a noticeable concentration of BDE 99 was detected in A. viridis (Table 2b) which might be due to coal combustion, urban sewage or the oil spill from pirate garages in the study areas (Kilunga et al., 2017).

Distribution of persistent organic pollutants (POPs) in soil
The concentrations of PCBs, PAHs, OCPs, DDTs and PBDEs in soils from eight garden areas were shown in Tables 3a and 3b. The PCBs congeners were measured in all soil samples and the sum of 12 PCBs (Σ12PCBs) concentrations ranged from 15.5 to 401.3 ng g −1 dw. The highest level was measured at the site RI with the value of 401.30 ng g −1 which is nearby N'djili River. Generally, concentrations of PCBs at the sites Mon, RI, CF, TS and LI were higher than those from the sites KI, SA and MO. Overall, the concentrations of PCBs were seemingly relatively higher in soils collected from urban impacted garden sites. In all soils, the predominant congeners were in order 101 > 153 > 118 > 138 > 52, and heptachlor biphenyl. These results were higher than those measured in urban rivers sediments in Congo DR (Kilunga et al., 2017). When comparing concentrations of PCBs in soils with recommended values, PCBs for most of the samples were found to be higher than probable effect level (676 mg kg −1 ), meaning that posed risks on the environment and gardening land. The dominating presence of higher chlorinated PCBs could be attributed to the local input sources and the impact of urban activities in receiving river ecosystems to the gardening land. Individual and total PAHs levels (ng g −1 ) in soil samples are presented in Table 3a. All analyzed PAHs congeners were detected in soil samples except the congener acenaphthylene. The concentration of total PAHs (Σ16PAHs) ranged between 93.5 and 469.0 ng g −1 in soils. In general, Σ16PAHs concentrations in soil samples were higher and more or less showed similar trend except  7.5 <7.5 <7.5 <7.5 <7.5 <7.5 <7.5 <7.5 <7.5 BDE153 7.5 <7.5 <7.5 <7.5 <7.5 <7.5 <7.5 <7.5 <7.5 for the sites MON and CF. The sum concentrations of (Σ16PAHs) measured in soil samples in this investigation show a significant increase in comparison to those previously reported in the Congo River (Mwanamoki et al., 2014;Verhaert et al., 2013). Among the sites, RI had the highest concentration of total PAHs, while the site MON showed the lowest level with the value of 93.5 ng g −1 . The result of this investigation showed that the total concentrations of upstream soil PAHs are relatively lower than that of downstream and PAHs levels increase downstream agricultural land. It was found that ΣPAHs of all soil samples did not exceed ERL, ERM and PEL. However, results show that the levels of Phenanthrene and Fluoranthene were higher than at the sites KI and CF, respectively. However, results show that the levels of Phenanthrene and Fluoranthene were higher than at the sites KI and CF, respectively. The result can be explained that these regions are associated with intensive anthropogenic activities and high road traffic. Therefore, it is crucial to monitor the levels of Phenanthrene and Fluoranthene in these sites at regular intervals. The PAHs sources were identified using several diagnostic ratios, as described previously (Budzinski et al. 1997, Yunker et al., 2002, Manneh et al., 2016. The results of diagnostic ratios are presented in Table S6 (SD). The ratios implied that petroleum combustion, and grass, wood, or coal combustion were the main sources of PAHs for most of the samples. The results of OCPs and DDTs levels are shown in Table 3b. It should be noted that the only pesticide detected in this study is chlorpyrifos-ethyl. The highest concentration was measured in Monastery, with a value of 23.05 ng g −1 . The reason may be attributed to intensive use of this pesticide on crops and runoff from rivers into gardening soils. Generally, the concentrations of DDTs were significantly higher in soil samples but lower than effect range median (ERM, 46.1 ng g −1 DW) (Feng et al., 2011). Remarkably, p,p′-DDE was detected in all samples. DDTs were the most dominant OCPs measured and it could be attributed to extensively use in agricultural and vector control in Congo DR.

Assessment of potential health risks
The estimated DI-PAH used to assess the human health risk of the selected PAH associated with consumption of PAH-contaminated A. viridis are given in Table S7 (SD). It appears that the DI-PAH values were not constant for the selected PAH and also highly varied with the sampling sites. DI-PAH of individual PAH values obtained for adults were ranged from 0.0001 to 0.00958 μg kg −1 body weight day −1 , 0.00016 to 0.01383 μg kg −1 body weight day −1 , 0.00005 to 0.0109 μg kg −1 body weight day −1 , 0.00001 to 0.003449 μg kg −1 body weight day −1 , 0.000089 to 0.007951 μg kg −1 body weight day −1 , 0.000118 to 0.010897 μg kg −1 body weight day −1 , 0.00003 to 0.001774 μg kg −1 body weight day −1 , 0.000059 to 0.004789 μg kg −1 body weight day −1 for KI, MO, SA, MON, RI, CE, TS and LI respectively. Compared with the virtually safe dose of 0.0005 μg kg −1 body weight day −1 set by the Dutch National Institute for Public Health and Environment (Khan and Cao, 2012), the daily intake of BaP through consumption of A. viridis for adults were ranged from 0.000256 (MON) to 0.00083 μg kg −1 body weight day −1 (SA). The sampling sites KI, SA, RI and LI indicate a daily intake of BaP relatively higher than the safe dose of 0.0005 μg kg −1 body weight day −1 . To estimate the carcinogenic potency for other PAH, the TEF values suggested for PAH by USEPA (2002) and BaP toxicological values were used. The TEQ values (ng g −1 ) calculated for the total PAH in A. viridis ranged from 4.8648 to 17.6318. The highest value (17.6318 ng g −1 ) was found at SA sampling site while the lowest (4.8648 ng g −1 ) at TS sampling site. The potential cancer risk (ILCR) was calculated by taking into account the importance of PAHs bioaccumulation in human bodies via food intake. According to the USEPA standard (USEPA, 1993(USEPA, , 2002, an ILCR less than 10 −6 indicates acceptable or negligible risk, while an ILCR greater than 10 −4 represents the serious risk (Bahrami et al., 2019;Nie et al., 2014;Zhao et al., 2014;Xia et al., 2010;USEPA, 1993USEPA, , 2002. The ILCR for PAH dietary exposure calculated in this study were found in the following order CE (1.24 10 −5 ) > KI (1.18 · 10 −5 ) > SA (9.8 · 10 −6 ) > LI (9.5 · 10 −6 ) > MO (5.2 · 10 −6 ) > TS (4.6 · 10 −6 ) > MON (4.3 · 10 −6 ) > RI (1.21 · 10 −6 ). For all sites the ILCR are less the priority risk level of 10 −4 and relatively higher than the acceptable risk level of 10 −6 . Attention should be paid for CE and KI sites where ILCR indicates values around 10 −5 .

Correlation between parameters
Principle component analysis (PCA) was performed to illustrate the contributions of total PCBs, PAHs, and DDTs in water, soil and A. viridis (Fig. 2). Fig. 2A demonstrates the impact of each variable and the relative correlations between them. All of the variables seem to have relevant significant impact (9-17%) on the total variance with the exception of the total PAHs in water and soil (3-5%), and total PCBs in water (0%). The reason that the total PCBs in water have no impact is that it remained below the detection limit at all the sites. Fig. 2B illustrates a total variance of 65.1%. The samples seem to cluster according to their physical proximity. Cluster 2 is less changing compared to cluster 1 because RI, LI, and CE are on the same river. Table S8 (SD) demonstrates the significance of the correlations among the same variables used in PCA. The only significant observed correlation is between the total DDTs in water and soil suggesting similar application, transport, and retention methods. These correlation results suggest that DDTs are likely from similar sources, whereas other POPs are probably from different sources.  petrogenic processes, pyrolytic sources or from grass, wood and coal combustion. -OCPs, such as p,p′-DDD and p,p′-DDT were detected in irrigation water and A. viridis samples. These results suggest that the sites investigated have been exposed by agricultural misuse of DDTs. -DI-BaP through consumption of vegetables for adults were relatively higher than the safe dose for the sampling sites KI, SA, RI and LI. -The ILCR do not indicates a significant potential for carcinogenicity.

Conclusion
-The results from this study provide important information on A. viridis contamination by POPs and PAHs and recommend strongly the control of pesticide use in studied gardening areas. -Thus in Kinshasa, urban agriculture control programs for POPs and fertilizer are very important in order to protect the consumer health. -The authors also ascertain that need for future epidemiological studies among the residents in the study area to assess the cancer incidence, kidney disease, and endocrine disease possibly from the dietary exposure to POPs.
This research warrants that further study on POPs should be conducted at regular intervals in others consumable vegetables cultivating and selling in studied areas. Based on our results, the authors ascertained to implement suitable measures and efforts by policymakers to reduce these contaminants to improve the quality of this one of the most consumed and exported plant in order to minimize human risks.
The research presented in this paper represents the first report regarding the POPs contaminating vegetable in studied region. It provides baseline information not only on the extent of water, soil and vegetable contamination by POPs as well as evaluation of potential human health risk in a tropical area, but also represents useful tools which can be applied to the similar regions.
CRediT authorship contribution statement GNN, AL, PS, CKM, FB, JNK and JP conceived and designed research; GNN, DMM. Al S, AL and DG performed research: sampling and laboratory analysis; All authors analyzed data and wrote the paper; All authors have read, corrected and approved the manuscript before submission.

Declaration of competing interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.