Chemical risks in drinking water of inhabitants in the basin of the Tonle Sap Great Lake

Cambodia is a tropical country where agriculture is the predominant employment sector for the rural Cambodian population accounting for 57.6% of the 8.8 million-strong labor force and contributing about 32.1% of GDP in 2011 (NCCC 2013). In order to realize a long-term vision of Cambodia to become a high middle-income country by 2030 and high-income country by 2050, the Royal Government of Cambodia has prepared and adopted an Industrial Development Policy (IDP), transforming and modernizing industrial structure from labor intensive industry to a skill-driven one by 2025 (RGC 2015). A general population census of Cambodia 2019 showed that the total population of Cambodia was approximately 15.6 million with a GDP per capita of 1,561$ in 2018 (NIS 2020). The adult mortality rate was 2.5 per 1,000 for women and 4.1 per 1,000 for men (NIS & DGHIM 2011) while the infant mortality rate was 18 per 1,000 live births, 16 in urban areas, and 19 in rural areas (NIS 2020). The Cambodian government has planned that 100% of the rural population will have sustainable access to an improved water supply by 2025 (MRD 2011). By 2020, it has been reported that approximately 73.0% of households had access to clean drinking water, 83.7% in urban areas, and 66.0% in rural areas (NIS 2020). Developments in the Mekong river basin, which might degrade the quality of the water source by nutrient, heavy metal, and pesticide contaminations, were likely to impact the drinking water quality due to the capacities of the available water treatment processes (Vanny et al. 2015). The assessment of industrial pollution load in Phnom Penh using an industrial pollution projection system revealed that the basic metal sector generated the most water pollutants in years 1994–2014 (San et al. 2018). This study reported that the plastic product and the coke and petroleum refinery sectors were the greatest water polluters in terms of BOD of 111 tonnes (58.47%) and 45 tonnes (23.71%) in 2014, respectively. The basic metal sector contributed the most TSS to water pollution by releasing 10,749 tonnes (65.23%) in 2014 followed by the sporting and athletic goods sector with emissions of 4,880 tonnes (29.61%). The average annual load of total water pollutants was approximately 2,152 ± 1,083 tonnes year⁻¹ with an annual emission growth rate of 108 ± 195 tonnes year⁻¹ (11.11%) (San et al. 2018). A previous study on the potential contamination of trace elements in shallow Cambodian groundwater showed that groundwaters from Kandal and Kraite were enriched in As, Mn, Ba, and Fe (Phan et al. 2013). Moreover, the assessment of water quality and trace metal contaminations in Mondolkiri province in the northeastern part of Cambodia revealed that the mean concentrations of Fe and Mn in tube wells and channel water exceeded the Cambodian aesthetic guideline (Phan et al. 2019). High arsenic levels (up to 100 μg L⁻¹) were also found in shallow groundwaters from the Tonle Sap Lake and coastal regions of Cambodia (Sovann & Polya 2014). However, it was reported that the shallow aquifer, which was primarily contaminated with pollutants, was chemically less of a health risk than the deep aquifer (Bennett et al. 2010). A study on the status of metal levels and their potential sources of contamination in Southeast Asian rivers revealed that the dissolved arsenic concentrations in the Tonle Sap–Bassac rivers were about 2.5 times higher than their background concentrations, the average dissolved metal concentrations of world natural river systems (Chanpiwat & Sthiannopkao 2014). It was reported that the use of rain and surface water from rivers, ponds, and lakes might reduce chronic arsenic poisoning, but they were often supplied without verification of their chemical and microbiological qualities, especially in rural areas (Feldman et al. 2007). Because some colorless, odorless, and tasteless contaminants such as arsenic (As), fluoride (F⁻), and nitrate (⁠⁠) could not be perceived by householders in their drinking water sources, the only practical way to prevent people from consuming unsafe contamination levels was to test the water quality (Guppy & Shantz 2011). To date, a nationwide water quality monitoring program has not been conducted, thus only a little data are available for health risk assessment and remedial options. Although numerous studies on Cambodian drinking water sources and their health risks have been documented in the Mekong river basin (Phan et al. 2010, 2013), the northeastern highland area (Phan et al. 2019), and the coastal area (Phan et al. 2021, 2023; Sao et al. 2023), human health risks of drinking water in the basin of the Tonle Sap Great Lake are little known. The emerging water contaminants described by Kumar et al. (2022) are of great importance for water quality assessment and monitoring, but the present study focuses on conventional contaminants because of the limitation of resources and facilities of our local laboratories. In the present study, the qualities of drinking water sources and human health risks of chemical contaminants through the drinking water pathway of inhabitants residing in the basin of the Tonle Sap Great Lake, Cambodia were assessed. The specific objectives were to: (1) determine the physicochemical properties, as well as microbial and chemical qualities of different drinking water sources in the Tonle Sap Great Lake basin; (2) assess the quality of drinking water sources by comparing with the Cambodian drinking water quality standard (CDWQS) and WHO's drinking water quality guidelines; and (3) assess human health risks of chemical contaminants through the drinking water pathway of inhabitants in the basin areas of the Tonle Sap Great Lake.

Water sampling was conducted in each province of the study area. However, sample collection was dependent on our accessibility to water sources. Water samples from tube wells (n = 52), dug wells (n = 13), stored rainwater (n = 39), ponds/lakes (n = 19), and canals/rivers (n = 24) were collected. Concurrently, pipe water from the point of use in households (n = 45) was also collected after informed consent was obtained. Groundwater was collected from tube wells after 5 min pumping of the standing water in the tube. Dug well water was sampled using a bucket to bring water samples from the well to the ground. Canal/river and lake/pond water were collected using the grab sampling method at a depth of about 0–30 cm. Simultaneously, on-site measurements of water dissolved oxygen (DO) and temperature were measured by using Hanna HI 9147 meter (Hanna, Italy). pH and ORP were measured by using a Hanna pH/ORP meter HI98191 (Hanna, Italy). Electrical conductivity (EC), total dissolved solids (TDS), and salinity were measured by using a Hanna Conductivity/TDS/Salinity meter HI 98192 (Hanna, Italy). Turbidity was measured by a Hach DR850 colorimeter (Hach, USA). A composite sample was filled into three different sampling bottles. Fresh (unfiltered, unacidified) samples were analyzed for physicochemical properties, as well as free Cl₂, F⁻, ⁠, and ⁠. Samples with acidified, filtered (0.45 μm) water samples were analyzed for As, Cu, Cr, Cd, Fe, Mn, Pb, and Zn. The water sample filled in a sterilized bottle was used to test Escherichia coli and coliforms. All collected samples were kept in an ice box during field sampling and then were transferred to a refrigerator and stored at 4 °C until analysis.

The measurement of free Cl₂, ⁠, and concentrations was conducted by using DR 1900 (Hach, USA) with the respective reagents supplied by Hach. The concentrations of Cu, Cr, Cd, F⁻, Fe, Mn, Pb, and Zn were analyzed by colorimetric standard methods (APHA 2017). E. coli and coliforms in the water sample were quantified using a modified membrane filtration technique developed by the United States Environmental Protection Agency (USEPA 2002). A 100 mL of water sample is filtered through a 47 mm, 0.45-μm pore size cellulose ester membrane filter, and then, the filter is placed on a 5-mL plate of MacConkey agar. The plate is incubated at 37 °C for up to 24 h. Bacterial colonies that grow on the plate are inspected for the presence of E. coli and coliforms with pink and yellow color, respectively. The suspected colonies from McConkey agar were enriched in Tryptic Soy Broth (Sigma Aldrich, US), after which Chromocult^® Coliform agar (Merck, Germany) was used to confirm E. coli colonies with dark-blue to violet colonies. The concentration was reported as colony-forming units (CFU) per 100 mL of the water sample and all samples were processed in duplicate.

HI is a hazard index that indicates the aggregate risk/risk of mixed chemicals. The non-carcinogenic effects of mixed chemicals are considered to occur in circumstances where HI is greater than one.

All statistical data analyses were performed by SPSS for Windows (version 15.0). The Mann–Whitney U test was used to verify the significant differences in physicochemical properties between tube wells and dug wells and those between pipe water and rainwater. The Kruskall–Wallis test was applied to verify the significant differences in E. coli and coliform among all water sources. Significance was considered in circumstances where p < 0.05.

The physicochemical properties and microbiological qualities of water sources in the basin of the Tonle Sap Great Lake are presented in Table 2. The pH of tube wells ranged from 4.57 to 7.74 (6.19 ± 0.78) where approximately 61.5% of tube wells had a pH < 6.50. The mean pH of dug wells was 6.36 ± 0.54 where approximately 53.8% of dug wells had a pH < 6.50. It was consistent with the previous report that the pH of shallow groundwater in the Tonle Sap Lake was slightly acidic to circumneutral, ranging from 5.8 to 7.6 (Sovann & Polya 2014). The mean pH of pond/lake water was 6.91 ± 0.94 (ranged from 4.47 to 9.08) while that of canal/river water was 7.27 ± 0.45 (ranged from 6.24 to 8.26); approximately 31.6% of pond water and 8.3% of canal water had pH < 6.50 while 5.3% of pond water had pH > 8.50. The pH of ponds/lakes was in a wider range than the previous study which reported that the Tonle Sap Lake had a pH that ranged from 7.04 to 8.61 in the dry season while it ranged from 7.91 to 8.40 in the rainy season (Yoshikawa et al. 2020). Similarly, the pH of canals/rivers was in a wider range than the previously reported pH of the tributary rivers of the Tonle Sap Lake (ranged from 6.60 to 7.93) (Yoshikawa et al. 2020). The pH of pipe water ranged from 5.09 to 9.18 (7.18 ± 0.70) where approximately 15.6% water had pH < 6.50 and 2.2% had pH > 8.50. Stored rainwater had a pH that ranged from 6.34 to 10.74 (7.93 ± 0.92) where approximately 5.1% water had pH < 6.50 and 17.9% had pH > 8.50. The EC of tube wells ranged 22.6–2,503.3 μS cm⁻¹ while that of dug wells ranged from 24.1 to 646.9 μS cm⁻¹. Sovann & Polya (2014) reported that the EC of shallow groundwaters from provinces around the Tonle Sap Great Lake, Cambodia ranged from 121 to 3,070 μS cm⁻¹, which was comparable to the present study. Out of all water sources, tube wells had the highest mean TDS of 261.8 ± 296.1 mg L⁻¹, followed by dug wells which had a mean TDS of 129.4 ± 108.0 mg L⁻¹. Although the mean TDS of tube wells was lower than 800 mg L⁻¹, approximately 9.6% of tube wells had TDS greater than the CDWQS (MIME 2004). Stored rainwater had the lowest mean turbidity of 1.1 ± 2.7 NTU whereas canal/river had the highest mean turbidity of 97.3 ± 229.2 NTU. The turbidity of pipe water ranged from 0 to 21 NTU with a mean of 4.9 ± 6.6 NTU. Approximately 28.8% of tube wells, 69.2% of dug wells, 26.7% of pipe water, 2.6% of rainwater, 42.1% of ponds, and 79.2% of canals had turbidity greater than 5 NTU, exceeding CDWQS (MIME 2004). Although DO, temperature, pH, EC, TDS, salinity, and turbidity were significantly different among all types of water sources (Kruskall–Wallis test, p< 0.05), except turbidity, they did not significantly differ between tube wells and dug wells (Mann–Whitney U test, p > 0.05). Further comparison showed that pipe water and rainwater were not significantly different in DO, temperature, ORP, EC, TDS, and salinity (Mann–Whitney U test, p > 0.05), but pH and turbidity were significantly different (Mann–Whitney U test, p < 0.05).

The average daily dose of a single chemical for adults is presented in Table 4 and for children is presented in Table 5. It is reported that approximately 55% of urban and 16% of rural residents have access to a safely managed drinking water service at their homes (MRD 2019). Our field observation revealed that rural inhabitants, who had access to either a tube well or a dug well, have used them in combination with rainwater. The exposure frequency of rainwater is about six months (EF = 180 days year⁻¹) in the rainy season (May to November). Tube wells and/or dug wells were used for another six-month period (EF = 180 days year⁻¹) in the dry season. Inhabitants who had pipe water connected at home or had access to other surface water sources, such as ponds and/or canals, used them year-round (EF = 365 days year⁻¹). The average daily dose of As up to 6.5 × 10⁻³ mg kg⁻¹ day⁻¹ was found in the basin of the Tonle Sap Great Lake. Previous studies had reported that the average daily dose of As was 3 × 10⁻⁷–10.75 × 10⁻³ mg kg⁻¹ day⁻¹ in the Mekong river basin of Cambodia (Phan et al. 2013) and 1.1 × 10⁻³–4.3 × 10⁻³ mg kg⁻¹ day⁻¹ in the Ha Nam province of Vietnam (Nguyen et al. 2009). The average daily dose of As in the basin of the Tonle Sap Great Lake was lower than in the Mekong River basin, but it was higher than that of the Ha Nam province of Vietnam.

Six different types of water sources, including pipe water from the point of use in the basin of the Tonle Sap Great Lake of Cambodia, were investigated on their qualities, and human health risks of chemicals through the drinking water pathway were assessed. Analytical results revealed that E. coli and coliform were detected in all water sources whereas pond/lake had the highest mean logCFU of E. coli. The concentrations of As, Ba, Cl₂, Cr, Cu, Fe, Mn, and Zn were significantly different among all types of water sources in the present study area. As distribution in water sources in the basin of the Tonle Sap Great Lake was comparable to a previous study, but it was lower than that of the Mekong river basin. Elevated Pb concentrations were found in tube wells, pipe water, and stored rainwater, whereas elevated Cd was found in all types of water sources. Elevated F⁻ concentrations were found in dug wells whereas elevated concentrations were found in tube wells. The health risk assessment revealed that some inhabitants residing in the basin of the Tonle Sap Great Lake are exposed to chemical contaminants through the drinking water pathway. Children are at a higher risk of chemical mixtures in their drinking water than adults. This study suggests that a nationwide water quality monitoring program should be regularly conducted. Moreover, drinking water should be properly treated based on the qualities of their sources and the best available water treatment technologies should be applied to prevent health risks of chemical contaminants to all consumers. Further studies on emerging water contaminants in various drinking water sources are warranted to better characterize health risks in order to prevent and/or minimize public health concerns in Cambodia.

This study was supported by the Higher Education Improvement Project (HEIP) funded by the Cambodian Government (IDA Credit No. 6221-KH) through Ministry of Education, Youth and Sport and GIST Research Institute (GRI) grant funded by the Gwangju Institute of Science and Technology (GIST) in 2023.

Pipe water from the point of use in households was collected after informed consent was obtained.

All authors read and approved the content of the final manuscript.

K.P., V.S., and C.O.C. conceptualized and wrote the manuscript. T.S. and V.S. reviewed and revised the manuscript. K.P., S.H., S.P., and N.T. performed sample and data analyses and graphing. C.O.C. approved the submission of the manuscript.

This study was supported by the Higher Education Improvement Project (HEIP) funded by the Cambodian Government (IDA Credit No. 6221-KH) through the Ministry of Education, Youth, and Sport.

All data are presented in this manuscript.

The authors have no conflict of interest to declare.

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

The authors declare there is no conflict.