Abstract

Cambodia has made progress towards addressing Sustainable Development Goal 6.1 “By 2030, achieve universal and equitable access to safe and affordable drinking water for all”, but challenges remain in fully realizing this target. We begin this paper by reviewing current country-wide access to safe and affordable water and subsequently report on the results of water sampling done for E. coli in a peri-urban area north of Phnom Penh that was conducted between 2018 and 2020. The sampling examined E. coli levels in source waters, including rivers, ponds, a large lake/wetland, wells, rainwater harvesting systems, piped water, and bottled water. Sampling from household storage containers and in-home drinking cups also was done to assess the effects that handling practices might have on exposure to E. coli. We show that country-wide, as well as in the peri-urban study area, there has been increased access to piped water. Piped water and commercially-available bottled water (0.5–1.5 L PET bottles) had the lowest E. coli levels in our study area, although such bottled water is not an affordable alternative for many peri-urban families. Surface pond water and the Tonle Sap River contained the highest E. coli levels and would pose the greatest risk associated with direct consumption. Handling practices may impact drinking water quality, as a significant difference (p=0.2) was found in E. coli levels between samples taken from commercially-available 0.5–1.5 L PET bottles and from household cups into which the bottled water was poured. There also was a significant difference (p<0.05) in E. coli levels between piped water sampled directly from the tap and piped water stored in bulk household containers. The geometric mean concentration of E. coli in large, covered, traditional outdoor storage jars used for rainwater harvesting was nearly 10 times lower than the same type of jars that were not covered, although due to the small sample size and variability in the data, the difference was not significant (p=0.5). Despite the increasing availability of piped water service in the study area, we found a diversity of water source practices, including use of rainwater harvesting, surface water, 20 L bottled water, and wells. These source waters can be safe, but must be routinely monitored. The study illustrates the advantages of field-based testing for effective screening of E. coli in peri-urban areas.

1 Introduction

Bartram and Cairncross (2010) argued that hygiene, sanitation, and water are the forgotten foundations of health when compared to the “big three” (HIV/AIDS, tuberculosis, and malaria) focus of the international public health community. A number of studies have noted improvements globally in WASH (Water, Sanitation, and Hygiene) practices over the past 20 years, yet estimated WASH-attributable deaths and DALYs (disability-adjusted life year) from diarrhoeal diseases remain in the range of 829,000 and 49.8 million, respectively (Hutton and Chase 2016; Prüss-Ustün et al. 2019; WHO 2021). Daly and Harris (2022) reported that a statistically significant increase in annual exposure to E. coli can occur when individuals supplement their improved water source with unimproved water for just 2 days annually. For the region of Southeast Asia, Chakravarty et al. (2017) concluded 364,000 deaths every year are attributable to inadequate WASH.

Despite the progress made in reducing microbial contamination, Thompson et al. (2021) noted that the burden of disease in Cambodian children is predominantly associated with diarrheal diseases and drinking water is an important vector source. Waterborne pathogens, including cryptosporidium, amoeba, E. coli, and Giardia duodenalis, have been connected to poor child growth in Cambodia (Poirot et al. 2020) and can negatively impact education due to higher classroom absenteeism (Hunter et al. 2014). Apart from the Phnom Penh Water Supply Authority (PPWSA) and water kiosks (Dany et al. 2000; Vanny et al. 2015; Sevea 2017; WaterAid 2018; PPWSA 2020), much of the work on microbial management in peri-urban areas of Cambodia has focused on the efficacy of various household treatment systems, including ceramic filters (Brown et al. 2008; Brown and Sobsey 2010; Murphy et al. 2010; Brown and Sobsey 2012), sand filters (Murphy et al. 2010; Stauber et al. 2012), solar disinfection (SODIS) (McGuigan et al. 2012), and boiling (Brown and Sobsey 2012; Thomas et al. 2015). While these different treatments are capable of significantly reducing microbial contamination and associated waterborne illnesses, efficacy of the systems are influenced by various factors related to individual design (e.g., proper packing of the sand for sand filters, boiling practices; variability of solar treatment time), level of contamination in the source water, and maintenance practices (Clasen et al. 2008; Stauber et al. 2012; McGuigan et al. 2012; Keane et al. 2014; Thomas et al. 2015; Irvine et al. 2016). A number of studies also have indicated that post-treatment handling and storage methods may result in re-contamination of the treated water in Cambodia and elsewhere (Jensen et al. 2002; Clasen and Bastable 2003; Wright et al. 2004; Brown and Sobsey 2012; Holman and Brown 2014; Thomas et al. 2015; Cassivi et al. 2021). WASH programs have been introduced by NGOs, schools, and the Ministry of Rural Development (MRD) to improve potable water practices in rural and peri-urban Cambodia. More recently, online approaches have been trialled to enhance WASH delivery (Alfvegren et al. 2019).

Irvine et al. (2006) summarized sources of drinking water in Cambodia based on the national census for the year 2000 and reviewed challenges in delivering safe potable water, including arsenic contamination of groundwater and microbial contamination of surface water. Innovative solutions developed by a local NGO, Resource Development International, Cambodia, to address these potable water challenges, including provision of household ceramic filters and rainwater harvesting systems, also were discussed. Since 2006, Cambodia has experienced remarkable economic growth and rapid urbanization, particularly around the primate city of Phnom Penh (Percival and Waley 2012; Mom and Ongsomwang 2016; Mialhe et al. 2019). Given this development, we felt it was time to revisit the potable water situation in peri-urban Phnom Penh to document current delivery and use practices and better understand the lived water experience of the non-elite peri-urban residents. In undertaking this research, we also help to document progress towards the Sustainable Development Goal 6.1 and make recommendations regarding possible ways forward. As such, the objectives of our study are fourfold: i) document and determine drinking water sources and quality for a peri-urban area north of Phnom Penh; ii) assess water quality characteristics for peri-urban homes in the study area, with particular focus on how handling practices may impact water quality; iii) illustrate that a fixed laboratory approach may not be necessary for effectively screening E. coli in peri-urban areas; and iv) review the multi-functional role of a large, natural wetland, Boeng Kob Srov, in providing effective wastewater treatment, drinking water, livelihood, and a sense of home.

Crocker and Bartram (2014) examined drinking water monitoring practises in Cambodia, Colombia, India, Jordan, Peru, South Africa, and Uganda with respect to responsibility of key agencies, approaches/methodologies, and cost. It was concluded that standardization of monitoring approaches was generally absent, except for some larger water suppliers in urban areas, and that sample analysis generally relied on fixed laboratories. Crocker and Bartram (2014) suggested that there was potential to optimize monitoring programs, particularly for non-piped supplies, by implementing field-based testing. Our third objective, therefore, specifically explored the efficacy of implementing Crocker and Bartram’s (2014) recommendation in the context of peri-urban Cambodia.

2 Background to the potable water situation in Cambodia

Subsequent to the national census reported in 2000 that included a summary of drinking water sources (Irvine et al. 2006), the census has been conducted twice more, in 2008 and 2019 (Table 1). Table 1 shows that nationally, both urban and rural areas experienced an increase in the use of piped water, although there was a slight decline for the urban areas between 2008 and 2019. This decline may have been associated with the official extension of Phnom Penh’s municipal boundaries into peri-urban areas north and southeast of the city. Rural areas have exhibited a progressive increase in the use of tube wells, a temporal trend that is of concern for certain regions (primarily the Mekong River floodplains) which have unsafe levels of arsenic (Polya et al. 2005; Berg et al. 2007; Feldman et al. 2007; Rodriguez-Lado et al. 2008; Sthiannopkao et al. 2010; Sovann and Polya 2014; Murphy et al. 2018). Use of surface water as a primary potable water source declined between 2008 and 2019 in rural and urban areas. Bottled water and rainwater harvesting may have a surprisingly small share of source water (Table 1), but it must be kept in mind that these data represent primary sources of drinking water. Rainwater harvesting may be important in the rainy season, but would need to be supplemented in the dry season, unless large collection systems were available. Similarly, bottled water would be cost prohibitive for many as a primary source of drinking water.

Figure 1 summarizes the national data reported for the U.N. Sustainable Development Goal 6.1. “By 2030, achieve universal and equitable access to safe and affordable drinking water for all”, indicator 6.1.1., “Proportion of population using safely managed drinking water services”. Country-wide, both urban and rural populations show a slow, monotonic increase towards access to safe drinking water sources, although collectively the access still remains low. Clearly, the urban population has a much higher access rate than rural areas and data generally are consistent with the urban and rural population access to piped water (e.g., 51.6% of urban population having access to piped water in 2019, Table 1, and 15.7% of rural population having access, Table 1).

Table 1 Primary sources of drinking water (% distribution) for the country of Cambodia. data source: National Institute of Statistics 2019.
(https://www.nis.gov.kh/nis/Census2019/Provisional%20Population%20Census%202019_English_FINAL.pdf Accessed 12 December 2022).

Source	Urban			Rural
	1998	2008	2019	1998	2008	2019
Piped (into dwelling or yard)	26.8	56.8	51.6	1.5	4.4	15.7
Public tap/standpipe	0	0	5.7	0	0	2.8
Tube well/borehole	12.5	14.7	12.9	15.5	29.5	32.6
Protected well	0	3.8	2.2	0	5.5	5.9
Unprotected well	16.7	4.5	1.4	45.1	24.2	8.6
Protected spring	0	0	0.2	0	0	0.4
Unprotected spring	0	0	0.1	0	0	0.5
Rainwater harvesting	0	0.5	1.5	0	1.1	3.5
Surface water (river, stream, lake)	13.3	6.2	5.4	31.2	26.8	13.7
Bought water	29.2	13.0	0	4.1	7.1	0
Tanker truck/lake or cart with drum	0	0	8.8	0	0	10.3
Bottled water	0	0	9.8	0	0	5.4
Other	1.5	0.5	0.6	2.7	1.4	0.7
Number of households	364,581	506,579	1,328,501	1,797,505	2,311,058	2,224,520

Figure 1 Proportion of population using safely managed drinking water services for urban and rural areas in Cambodia, 2000–2017.
(data from https://unstats.un.org/sdgs/indicators/database/ Accessed 26 August 2021).

3 Methods

3.1 Study Area

Irvine et al. (2006) discussed the physical characteristics of Cambodia in more detail, but briefly, the country has a tropical monsoon (Am) climate based on the Koppen classification scheme, with two distinct seasons, a rainy season between May and October, and a dry season from November through April. Mean annual precipitation is 1,840 mm, with 80–90% falling in the rainy season. January can be particularly dry, with a monthly average of only 14 mm for the period 1991–2020 (https://climateknowledgeportal.worldbank.org/country/cambodia/climate-data-historical Accessed 7 November 2021). Monthly mean temperature for the same period of record, 1991–2020, ranged between a high of 29.4^oC in April, and a low of 25.8^oC in December (https://climateknowledgeportal.worldbank.org/country/cambodia/climate-data-historical Accessed 7 November 2021). Although there is a distinct cooler season between November and February, in general, temperatures remain relatively warm throughout the year, with the lowest mean minimum temperature of 20.6^oC occurring in January. The country exhibits some spatial variability of rainfall (Tsujimoto et al. 2018) and temperature related to elevation and relative proximity to the coastline.

Sampling for E. coli analysis was done in the Kob Srov watershed and vicinity, approximately 18 km northwest of the Phnom Penh Central Business District (Figure 2). The area is home to a number of villages that line the Tonle Sap River (and including floating houses), as well as within the general watershed for Boeng Kob Srov. The area’s proximity to Phnom Penh is such that higher-end housing development is occurring, with people commuting daily into the city for work. Extensive agricultural practice still exists throughout the watershed, with crops including morning glory, lotus, and rice. A large area of the eastern watershed is inundated during the rainy season (Figure 3). Potable water sources in the Kob Srov watershed are quite varied and include surface ponds, groundwater wells, rainwater harvesting systems, the Tonle Sap River, and Boeng Kob Srov (also known as Boeng Tamouk).

Figure 2 Cambodia (top left), and municipality of Phnom Penh (top right). Sample sites, Phases I, II, and III (lower).

The green pentagons represent the routine sample locations from Phase I, the yellow circles represent the Phase II sample village locations, while the brown circles represent the sample village locations from Phase III. The communes in the Phase III sampling have been labelled and shaded. 

Figure 3 Flooded Kob Srov watershed along the Tonle Sap River, during the rainy season, September 2019. Photo by authors.

Sahmakum Teang Tnaut (STT) (2021) noted that Boeng Kob Srov is the largest of the wetlands surrounding Phnom Penh, with a surface area of 3,240 ha. STT (2021) also reported that approximately 300 families (with a total of 1,000 people) live on the shores around the lake, often occupying informal housing, and earning a living through fishing, aquaculture farming, and home-based businesses. Boeng Kob Srov provides the additional ecosystem service function of treating wastewater from the northern area of Phnom Penh, much as Boeng Cheung Ek does for the southern area of Phnom Penh (Visoth et al. 2010; Im et al. 2015; Irvine et al. 2015; Somara and Mihara 2021).

3.2 E. Coli Sampling

The sampling was done in three phases, representing different thematic strands:

Phase I

Boeng Kob Srov, the Tonle Sap River, and wastewater discharges from north Phnom Penh to Kob Srov, were sampled and analyzed for E. coli on 12 occasions between 4 May 2018 and 2 January 2020, covering both the rainy season and dry season. This routine sampling at master stations was conducted over the course of the entire study period to document general baseline surface water conditions and the relative efficacy of Boeng Kob Srov in treating wastewater discharges. A total of 156 regular samples were collected and for QA/QC purposes, duplicates were collected at three randomly selected sites on three different dates. As noted in earlier research (Irvine et al. 2006; 2008; 2015; Vuong et al. 2007; Visoth et al. 2010), Phnom Penh does not have conventional wastewater plants to treat municipal and industrial discharges, but instead has relied on a ring of natural wetlands around the city to provide this service. Research on the efficacy of wetland treatment, to date, has focused on Boeng Cheung Ek to the south of the city, and Boeng Kak, which was more centrally located. Boeng Kak has now been entirely infilled for development and existing plans for Boeng Cheung Ek would see its area substantially reduced (Schneider 2011; Doyle 2012; Irvine et al. 2015; Loc et al. 2020; Irvine et al. 2021). To our knowledge, this is the first report on wastewater quality entering Boeng Kob Srov, and the treatment efficacy of the wetland, although Eliyan et al. (2022) recently estimated phosphorus and nitrogen loading rates to the wetland from fecal sludge disposal practices.

Phase II

Drinking water was sampled from home storage containers in the case of rainwater harvesting or piped water sources, and directly from the bottle in the case of commercially purchased (bottled) water between 5 and 13 December 2018. A total of 53 samples were collected from 53 households (1 sample per household) and analyzed for E. coli. The households represented 6 communities: one being predominantly ethnic Cham people (Cambodian Muslim, historically of Malay-Polynesian origin), one a predominantly Vietnamese village, and four predominantly Khmer villages.

Phase III

The three stages in the drinking water train were examined: i) drinking water sources, including wells, surface water (ponds, lakes, and rivers), commercially available bottled water, and piped water to individual homes; ii) household bulk storage containers situated outside of the house; and iii) point-of-use inside the home. Household bulk storage refers to water withdrawn from a source which is subsequently stored in large containers such as traditional ceramic water jars or cement water tanks situated outside of the house. In this sampling, we also included rainwater harvesting systems. The water from household bulk storage may be used for domestic purposes (cooking, washing) in addition to drinking. In contrast, point-of-use sampling was done from cups, plastic bottles, or other small storage containers inside the home from which the water normally was consumed by the family. It also should be noted that the sampled source waters (e.g., wells, surface water bodies) may be the source for more than one sampled household. Sampling was conducted between 9 May 2019 and 28 July 2019, with a total of 117 samples collected for E. coli analysis in ten communes. Between 1 and 5 villages were sampled in each commune. The sampled villages predominantly were Khmer, although one village was predominantly Vietnamese, and one was predominantly Cham. Of the 117 samples collected, 45 samples were obtained from sources, 41 samples from household bulk storage, and 31 samples were point-of-use.

Convenience sampling was used for the selection of households in Phases II and III. Phase III was differentiated from Phase II, as the sampling was done with a refined sampling strategy that more explicitly addressed the potable water chain from source to point-of-use consumption. The locations of the Phase I routine monitoring stations, as well as the locations of the villages included in the Phase II and Phase III sampling, are shown in Figure 2.

In all cases, the Coliscan Easygel system from Micrology Labs, Goshen, Indiana, USA, (https://www.micrologylabs.com/water/ Accessed 23 Aug 2021) was used for the E. coli analysis. Each water sample was collected with a new, sealed, sterile plastic pipette. For Phase I, all samples were collected directly from the water source by the pipette, with the exception of the sewage samples (Figure 4). Due to safety concerns, this direct sampling was not possible for the sewage discharges, and therefore we used a newly purchased and then emptied commercial PET drinking water bottle, attached to a bamboo rod. The plastic bottle was rinsed 3 times with the sewage before the analyzed sample was collected. Given the high levels of E. coli in the sewage, we do not anticipate contamination with this procedure (Figure 4). In Phase II, the piped water was retrieved either from a household container used to store the water or from a household drinking glass provided by the homeowner. Commercially-available bottled water (either 0.5–1.5 L PET or 20 L “blue bottles”) were sampled directly with the pipette, as were the traditional ceramic water jars. In Phase III, the piped water was collected directly from the spigot in a clean commercially obtained PET drinking water bottle, and a similar procedure was used at the well-head of tube wells. The PET bottles were rinsed 3 times with the piped or well water prior to final sample collection. The on-site bucket was used to collect samples from shallow dug wells. Surface water sources and all large bulk storage containers were sampled directly with the pipette (Figure 4), as was the case with the point-of-use sampling.

Figure 4 Sampling the Tonle Sap River from a floating house directly with a sterile pipette (upper left), from traditional ceramic water jars directly with a sterile pipette (upper right), and sewage discharge with a clean PET bottle attached to a bamboo rod. Photo by authors.

The volume of water withdrawn by each sterile pipette depended on the water source, with 1 mL being used for surface water sources; 5 mL for wells, drinking water storage containers, point-of-use sampling, and commercial water bottles; and 0.037 mL for the sewage discharges. The sterile pipette was used to dispense the water sample into the Coliscan Easygel growth medium contained within individual plastic vials. The dispensing from the pipette into the growth medium vial occurred on-site. The vials were gently swirled to distribute the inoculum and then placed in a cooler for transport to laboratory space at the Royal University of Phnom Penh for plating and counting. Hold times were 6–8 hours, which is not excessive and should help to ensure quality results (Paretti et al. 2018). The plated samples were incubated at room temperature for 48 hours and all purple colonies were then counted as E. coli. This procedure followed the manufacturer’s methodology and further detail can be found at https://www.micrologylabs.com/coliscan-easygel/ (Accessed 7 August 2023). In a comparison between Coliscan tests and standard membrane filtration (done by a certified public health laboratory), Irvine et al. (2011) reported a correlation of 0.98 (n=21) over a range of 0 to 120,000 cfu/100 mL. The Coliscan Easygel system has been used successfully at other locations in Southeast Asia (Visoth et al. 2010; Irvine et al. 2013; Irvine et al. 2016; Bhowmick et al. 2017) and under a wide range of wastewater, natural water, and drinking water conditions for countries from the global north and global south (Elmore et al. 2005; Higgins et al. 2005; Cui et al. 2011; Wang et al. 2015; Rogers-Brown et al. 2016; McFadden et al. 2017).

The Mann-Whitney U nonparametric test using raw data rather than log-transformed data was used to assess differences between samples, although per standard reporting practices for E. coli we also present the geometric mean values. The Mann-Whitney U tests were conducted using XLSTAT (https://www.xlstat.com/en/ Accessed 18 August 2023).

4 Results and discussion

4.1 Phase I – Routine sampling at master surface water stations

The master sample locations are shown in Figure 5 and the geometric mean E. coli results for the sites are plotted in Figure 6. There are four important trends that can be observed. First, it is not surprising that the E. coli levels in the sewage discharges from Phnom Penh (Figure 5; S3 and S9, Figure 7, top) are relatively high, although we cannot conclusively determine why S3 levels are greater than S9. The difference in E. coli level at these two sites may be due to differences in source strengths (e.g., domestic vs. industrial), but also the canal leading to S9 tends to be heavily clogged with water hyacinth for approximately 2 km. These macrophytes may serve to filter the wastewater. In comparison, geometric mean E. coli levels in wastewater discharging from south Phnom Penh to the Boeng Cheung Ek wetland were reported in the range of 396,00–5,632,000 cfu/100 mL (Visoth et al. 2010; Sovann et al. 2015). An earlier study conducted by JICA (1999) found fecal coliform levels at six sewer locations throughout Phnom Penh ranged between 3,600 and 4,600,000 cfu per 100 mL. In general, E. coli levels for sewage samples from this study are consistent with those reported for other locations in Phnom Penh.

Figure 5 Master sample station locations.

Figure 6 Geometric mean E. coli levels at master stations sampled in study Phase I.
For calculation of the geometric mean, values of 0 cfu/100 mL were replaced with a value of 0.01 cfu/100 mL. Sample size, n=12 for S1, S4, S5, S6, S7; and n=6 for S3 and S9.

Second, based on the results from S4, S5, and S6 (Figure 7 lower left), Boeng Kob Srov is effective in treating north Phnom Penh’s wastewater. The geometric mean levels are slightly higher at S4, which is closest to the S3 discharge, lower at S5, located on the wastewater discharge side of an earthen dike that divides the boeng, and lower still at S6 which is near S5 but on the opposite side of the dike and more hydraulically isolated from the sewage discharges (Figures 5 and 7). The Mann-Whitney U test indicated that E. coli levels were significantly different (p <0.05) for sites S4, S5, and S6, as compared to the sewage discharge at S3. Comparing the geometric mean E. coli level for the wastewater at S3 with those from S4, S5, and S6, the reduction was 99.9858%, 99.9962%, and 99.9996%, respectively. Irvine et al. (2008) reported E. coli levels were reduced by 99.9% within 200 m from the point of entry to the Boeng Cheung Ek wetland in south Phnom Penh. The distance between site S4 and the wastewater discharge at S3 was 900 m. The Boeng Cheung Ek wetland treatment was aided by mechanical filtration associated with a dense mat of floating water spinach grown as part of active water-based agriculture, while Boeng Kob Srov has a larger dilution potential than Boeng Cheung Ek due to its size.

Figure 7 Master sample sites S9 (top); S6 (lower left); S1 (lower middle); S7 (lower right). Photos by authors.

Third, water leaving Boeng Kob Srov discharges to the Tonle Sap River via a 1.7 km canal that is lined with light industry, informal housing, and small wet markets. The geometric mean E. coli at S1, near the mouth of this canal (Figure 7, lower middle) exhibits a considerable increase in E. coli levels due to untreated discharges from these sources. The Mann-Whitney U test also indicated that the E. coli levels were significantly different at S1 compared to S5 (p <0.05). The flow from this canal enters the Tonle Sap River approximately 3.5 km upstream of site S7. Site S7 (Figure 7, lower right) was located on the Tonle Sap River in an area where both floating houses and villages line the banks, and in particular, is a source of domestic water for the floating households. Based on sampling done near the water intakes on the Tonle Sap River (7–8 km downstream of site S7), Dany et al. (2000) reported E. coli levels of around 20,000 cfu/100 mL, which is consistent with the results at site S7 (Figure 6).

Finally, the surface water site results were categorized by season (wet vs. dry) and results are summarized in Table 2. The sample size, particularly for the dry season, is small and results should be viewed with caution, but the spatial trend in E. coli levels appears to be stronger than the temporal (seasonal) trend. For example, Mann-Whitney U tests indicated no difference (p >0.5) in E. coli levels between seasons at S1, S4, S5, and S7. The Mann-Whitney U test did, however, suggest there was a significant difference (p <0.05) between the rainy season and dry season E. coli levels at site S6, with the rainy season being higher. Nakhle et al. (2021) reported higher E. coli levels occurred in the Lao PDR section of the Mekong River and tributaries, noting that there was a strong correlation with total suspended solids levels, which suggested increased runoff, and erosion of soils and river channels were an important driver.

The results of the duplicate sampling are presented in Table 3, together with the Relative Percent Difference (RPD). There is good correspondence between the pairs of duplicate samples, with the RPD ranging between 11.4% and 14.8%. The state of California (https://www.waterboards.ca.gov/water_issues/programs/tmdl/records/region_1/2013/ref4112.pdf Accessed 7 August 2023) recommends that for surface waters, the RPD for E. coli should not exceed 20%, while Nowicki et al. (2021) reported an RPD of 25.6% for groundwater samples that were potable water sources.

Table 2 Geometric mean E. coli Levels (cfu/100 mL¹) for master stations, by season.

Site	Rainy Season, cfu/100 mL, n=8	Dry Season, cfu/100 mL, n=4
S1	49,802	53,686
S4	357	42
S5	20	259
S6	15	1
S7	19,503	15,772

¹For calculation of the geometric mean, values of 0 cfu/100 mL were replaced with a value of 0.01 cfu/100 mL.

Table 3 Results for duplicate samples.

Site	Sample, cfu/100 mL)	Duplicate sample, cfu/100 mL	Relative percent difference
S7	13,000	11,500	12.2%
S7	92,600	107,400	14.8%
S9	3,515,400	3,939,300	11.4%

4.2 Phase II - Village drinking water

E. coli levels varied by source water and the data are summarized in Table 4. Piped water and bottled water generally were reliable and clean options as a drinking water source. However, some special cases with respect to the results in Table 4 also are worth noting. In particular, the two highest E. coli results for piped water (19,200 and 1,140 cfu/100 mL) were sampled from uncovered, outdoor, traditional water jars located adjacent to a busy road, and from an open, plastic storage container at the household (Figure 8). In the case of the uncovered outdoor traditional water jars, the household purchased piped water from a neighbour across the road because the piped system had not been extended to this particular household. Sample results from these two households suggested that handling and storage methods of the drinking water may be a critical factor with respect to contamination risk and as such this issue was examined more specifically in Phase III of the study. E. coli levels in the rainwater harvesting samples were slightly higher than the other samples, but with proper treatment prior to use, should not present a health risk.

Table 4 Summary statistics of E. coli levels (cfu/100 mL) by source, Phase II.

	Piped water	20 L Commercial bottled water	Commercial bottled water, 0.5-1.5 L PET bottles	Rainwater
n	16	29	2	4
Geometric mean1	0.2	0.03	0	20.2
Min	0	0	0	0
Max	19,200	800	0	580
% with 0 cfu values	75%	86%	100%	25%

¹For calculation of the geometric mean, values of 0 cfu/100 mL were replaced with a value of 0.01 cfu/100 mL.

Figure 8 Traditional, uncovered water jars storing piped water but having elevated E. coli levels (left); Piped water stored in a plastic bucket at the home exhibiting elevated E. coli levels (right).

Discussions with the 16 community members who indicated their households used piped water revealed that 53% received the water from the municipality, while 47% received the piped water from private companies. Of this same group, 18% indicated that they allowed the piped water to stand in a storage container for several days prior to use because this helped to clear a noted smell, and one family reported they experienced rashes after showering with the piped water.

4.3 Phase III - Village drinking water from source to end-of-cup

E. coli levels varied by source water and the data are summarized in Table 5. Surface pond water and the Tonle Sap River represent the highest E. coli levels and would pose the greatest risk associated with direct consumption. The post-Khmer Rouge recovery philosophy for potable water management was to minimize risk of biological contamination through migration from surface water to well water as the primary source in rural and peri-urban areas. The geometric mean E. coli levels from the wells in this study were lower than the surface waters (river or pond), although one household well exhibited elevated levels at 34,500 cfu/100 mL. E. coli contamination was limited in the piped water, with 2 of 14 samples having levels of 60 and 960 cfu/100 mL, and the remaining samples being 0 cfu/100 mL. Similarly, only one (of 13) bottled water samples had a positive result for E. coli (140 cfu/100 mL).

Table 5 Summary statistics of E. coli levels (cfu/100 mL) by source, Phase III.

	Piped Water1	Pond2	Well3	Rainwater (covered jar)	Rainwater (jar not covered)	River4	Bottled5
n	14	3	9	16	5	6	13
Geometric mean6	0.03	80,142	3.7	65	569	64,929	0.02
Min	0	51,300	0	0	0	11,900	0
Max	960	136,700	34,500	31,040	26,920	188,800	140

¹Piped water sampled at 14 individual homes; ²Surface water from 3 ponds sampled, but 3 ponds serviced 17 surveyed households; ³Combination of community wells (n=2) that serviced 11 surveyed households and 7 household wells owned by a single family; ⁴Samples do not include Phase I master station, but represent other locations on the Tonle Sap River; ⁵Samples were from the 20 L “blue bottles” (n=11) and 2 from smaller 500 mL store-bought PET bottles; ⁶For calculation of the geometric mean, values of 0 cfu/100 mL were replaced with a value of 0.01 cfu/100 mL.

When we tracked the piped water from source to bulk household storage, the geometric mean concentration increased from 0.03 cfu/100 mL to 332 cfu/100 mL. A Mann-Whitney U test showed that the E. coli levels for the piped water source and bulk storage water were significantly different (p <0.05). The commercially-bottled water exhibited E. coli levels of 0 cfu/100 mL for 92% of the samples analyzed, while 62% of the at-the-cup (point-of-use) samples had E. coli levels of 0 cfu/100 mL, with 2 at-the-cup (point-of-use) samples having excessive E. coli levels, in the 24,720–58,960 E. coli/100 mL range. A Mann-Whitney U test also indicated that the E. coli levels for the bottled water and at-the-cup were significantly different, but only at p=0.2.

The geometric mean concentration of E. coli in traditional water jars used for rainwater harvesting that were observed to be covered and reported as being routinely covered (65 cfu/100 mL) was nearly 10 times lower than traditional water jars that were not routinely covered (569 cfu/100 mL), although due to the small sample size and large standard deviations, a Mann-Whitney U test could not detect a difference between the two sample types (p=0.5). We also note here the results in Phase II of the study which showed the two highest E. coli results for piped water were collected from uncovered, outdoor traditional water jars located adjacent to a busy road, and from an open, plastic storage container at the household. Some studies have reported a significant difference in bacterial contamination between covered and non-covered water jars (Wright et al. 2004; Shaheed et al. 2014), while others have not found a difference (Trevett et al. 2004; Vannavong et al. 2018). Although the weight of evidence in this study suggests covering water storage containers may be helpful in reducing bacterial contamination, the issue should be examined further. Nonetheless, as an interim, simple, precautionary approach, we recommend jars be covered.

There were some important differences in the source water between Phase II and Phase III samples. In particular, the 20 L commercial bottled water was observed much less frequently in Phase III compared to Phase II, while no wells were recorded as a source water in Phase II. Access to piped water expanded over the course of the entire study. For example, one of the villages that did not have piped water in Phase II had full access to piped water in Phase III of the study. Finally, as a way of linking Phase II and Phase III, and also as a measure of QA/QC, we re-sampled the piped water for a subset of five houses along the same canal connecting Boeng Kob Srov and the Tonle Sap River (see Figure 7, lower middle). The samples were collected 9 months apart (8 December 2018 for Phase II and 21 September 2019 for Phase III). E. coli for the piped water samples at all five houses in Phase II was 0 cfu/100 mL. For Phase III, E. coli was 0 cfu/100 mL at four of the houses, but was 960 cfu/100 mL at one of the houses. The head of household at this site noted there had been recent foul/wastewater odours associated with the piped water. Furthermore, the Phase III sampling was conducted during the rainy season and an inspection suggested there may be some ingress from the canal into the household piping system at this location.

4.4 Lessons learned

Piped and bottled water generally were the safest sources of potable water for the Kob Srov study area. We noted the expansion of the piped water network within our peri-urban study area, even between Phase II and Phase III, which clearly was related to the urban expansion of Phnom Penh and associated socio-economic drivers. The piped water suppliers were a mix of municipality and private companies, consistent with the trends identified for other areas of Cambodia (Sevea 2017). The World Bank (2016) reported that apart from Myanmar, Cambodia had the lowest access to piped water supply in Southeast Asia and this would be particularly true for the rural population (Table 1). WaterAid (2018) noted that “Over the past decade, many of Cambodia’s 12 million rural residents have experienced changes to their water supply practices…”  and that “Approximately 27% of rural households now purchase water from a service provider – most commonly from a piped water supply system or a bottled water distributor – compared to just 11% in 2009.” This trend also is generally observed in Table 1. Water Service Providers (WSPs) in urban areas fall under the purview of the Ministry of Industry and Handicraft (and within the Department of Potable Water Supply) while jurisdiction in rural and peri-urban areas is assigned to the Ministry of Rural Development (and within the Department of Rural Water Supply). In addition, the Cambodia Water Supply Association (CWA) began operation in 2012 through the financial and technical support of USAID. As an independent, non-profit and non-political organization, it acts as a resource center to WSPs. Currently, it has 247 WSP members operating in 23 provinces of Cambodia (http://www.cwa.org.kh/about-us/ Accessed 7 November 2021). WaterAid (2018) reported a total 13 public urban water utilities exist in Cambodia and an estimated 530 WSPs in urban and rural areas are known to the relevant government agencies, but Sevea (2017) found that not all WSPs are licensed and the water quality of unlicensed WSPs can be suspect. The Ministry of Rural Development recently announced the goal of “…providing 100 per cent clean water supply and sanitation services throughout the Kingdom by 2025…” (https://www.phnompenhpost.com/national/100-access-clean-water-nationwide-2025 Accessed 8 January 2023). While progress has been made, as shown in Figure 1 and from the results of this study, and the goal is laudable, we are sceptical that such an ambitious target can be met in two years.

Hygiene practices at the household level are a concern, such that even if the source water shows no contamination from E. coli, unsanitary storage and handling can result in contamination. This risk can be managed through public awareness campaigns that address proper treatment of water, safe storage, and safe handling practices, including cleaning the household storage containers and covering rainwater jars with lids. Some progress has been made under various WASH programs in Cambodia with respect to in-home water handling (Murphy et al. 2009; Orgill et al. 2013; Hunter et al. 2014; Alfvegren et al. 2019; Nhim and Mcloughlin 2022), but further work with respect to training the trainers and program implementation is needed.

Despite the prevalence of piped water service in the study area, we also found a diversity of water source practices, with a considerable reliance on rainwater harvesting, in addition to surface water, 20 L bottled water, and wells. Indeed, some quite extensive and sophisticated household rainwater harvesting systems were observed in the study area during the Phase III sampling (Figure 9). The diversity of water sources and practices can help to increase community resilience to a changing climate.

Figure 9 Extensive roof and rainwater harvesting system, including covered traditional ceramic water jars.

Some households in both Phase II and Phase III of the study reported concerns about odour and rashes associated with the piped water and these families generally allowed the water to stand in storage for a couple of days to clear the odour. We did not conduct measurements, but we suspect that the odour (and rashes) was due to excessive chlorination, which volatized while the water was standing. More recently, sampling was conducted by a Royal University of Phnom Penh Environmental Science Department class to test chlorine residual in piped water at homes within the study area and it was confirmed that chlorine residual exists in levels consistent with dosage at the water source. This issue should be considered further in a follow up study to optimize chlorine dosing, thereby accruing both environmental and household use/perception benefits.

The Coliscan Easygel system is robust and easily implemented in the field. Although we plated all samples in a laboratory at the Royal University of Phnom Penh, incubation is done at room temperature and does not require specialized equipment or highly-skilled staff for analysis. As such, this system is adaptable for use in any rural area having a clean space and would be an option for implementing field-based testing, following the suggestions of Crocker and Bartram (2014). As noted, Irvine et al. (2011) found that the Coliscan Easygel results matched well with those from a certified public health laboratory in the United States.

Finally, our study has documented complex surface water interactions, including the movement of urban wastewater through natural wetland areas and cycling back to the Tonle Sap River upstream of Phnom Penh. Throughout the cycling process, we observed the natural cleansing of biologically-contaminated water as well as the re-contamination from both urban and peri-urban sources. This treatment process is an ecosystem service co-benefit provided by Boeng Kob Srov, in addition to other ecosystem services such as food security and livelihoods, runoff regulation, carbon sequestration, mitigation of urban heat island effect, and cultural services such as home and sense of place. Sadly, the natural wetland areas that ringed Phnom Penh and provided these multiple ecosystem services (e.g., Boeng Cheung Ek, Boeng Kak, Boeng Kob Srov) are disappearing to accommodate development of satellite cities and other infrastructure tributary to Phnom Penh (Schneider 2011; Doyle 2012; Irvine et al. 2015; 2021; Loc et al. 2020; Ro et al. 2020), which truly is a tragedy of the commons. STT (2021) thoroughly documented the allocations of Boeng Kob Srov surface area between October 2018 and January 2021, to individual wealthy citizens, government agencies, and 23 families to be relocated from other areas of the wetland. These allocations eventually will result in an infilling of 756 ha (23.3% of the wetland area) and STT (2021) noted that in the official Phnom Penh municipal master plan it is expected only 2,140 ha (66%) of the wetland will remain by 2035. Perhaps in an ironic twist, part of the recent infilling of Boeng Kob Srov is for a 75 ha Phnom Penh National Park, intended to make the area more attractive, liveable, and to preserve rare trees for future generations (https://www.khmertimeskh.com/50824736/phase-one-of-phnom-penh-national-park-expected-to-be-completed-next-month/ Accessed 10 December 2022).

5 Conclusion

As illustrated by the National Institute of Statistics data and confirmed by more localized data collected in our field study, Cambodia has made progress towards addressing Sustainable Development Goal 6.1. “By 2030, achieve universal and equitable access to safe and affordable drinking water for all”, certainly since our earlier work published more than 15 years ago (Irvine et al. 2006). In particular, there has been expanded availability of higher quality piped water in peri-urban areas, which has been driven by improved socioeconomic conditions over the past decade. However, we also noted that in our study area, a considerable diversity of drinking water sources still exists, in part because of proximity to large water bodies (Tonle Sap River, Boeng Kob Srov), but also rather elaborate rainwater harvesting systems had been constructed. These diverse sources can enhance water resiliency, but appropriate construction, maintenance, and handling must be practiced. First-flush diversion or traps should be considered in the rainwater harvesting system design. Eaves, downspouts, and collection jars should be regularly cleaned and installation of coarse screens on the eaves may reduce contamination from birds.  

Considerable research and applied work have been undertaken by the academic community, NGOs, and government agencies in relation to potable water quality in Cambodia over the past 15 years and this has increased citizen awareness of the issues. However, we found that bulk storage significantly increased E. coli levels of pipe-sourced water and point-of-use samples had significantly higher E. coli than commercially-purchased bottled water. Furthermore, the weight-of-evidence from the study suggests covering traditional water jars may reduce potential for contamination, but a larger study is needed to confirm this statistically. We suggest that more work needs to be pursued, particularly with respect to water handling and education, for Cambodia to continue its path in meeting Sustainable Development Goal 6.1.

Wastewater discharges were effectively treated by Boeng Kob Srov. However, there was considerable spatial variability in E. coli levels for this surface water system, with lower levels moving away from the wastewater discharge points, into the wetland, but recontamination from local sources occurring along the discharge canal to the Tonle Sap River. This type of surface water system connectivity needs to be considered in managing downstream water intakes. Given the high E. coli levels in the Tonle Sap River, households using it directly as a source of water should practice thorough treatment prior to consumption.

The Coliscan Easygel system proved to be a flexible and reliable approach to monitoring E. coli levels in a variety of source waters. This type of system may be helpful, particularly in more remote areas of Cambodia, by providing quantitative, field-based monitoring, following the recommendations made by Crocker and Bartram (2014). The approach is cost-effective and does not require specialized labs, equipment, or highly trained technicians. The system is capable of accommodating a large range of E. coli levels in different water types, from 0 cfu/100 mL in drinking water to millions of cfu/100 mL in wastewater. Although our duplicate sample program was limited in scope, the RPD between duplicates was in the 11.4% to 14.8% range, which is acceptable. The Coliscan system also showed consistent results of 0 cfu/100 mL for piped water sources at 4 of 5 households sampled nine months apart. The E. coli level at the fifth house was 960 cfu/100 mL in the second round of sampling and the piped system here seems to have experienced some local contaminated water ingress. This admittedly limited sampling of 5 households nonetheless illustrates the robustness of the Coliscan system and the value of longitudinal monitoring to identify and correct contamination problems, with the goal to reduce incidence of waterborne disease. 

Acknowledgments and declarations

The authors would like to thank Bong Phors Thong for his outstanding assistance in the field over the past 20 years and including all three phases of this study. We also would like to thank the four reviewers who provided thoughtful and constructive comments that served to improve the manuscript. All research procedures involving human participants were done in accordance with the ethical standards of the 1964 Helsinki declaration and its later amendments or comparable ethical standards. Informed consent was obtained for all research components that involved human participants and was in accordance with standard ethical practice.

References

1. Alfvegren, F., B.S. McIntosh, and V. Chan. 2019. “Development of an e‐Learning Course to Disseminate Guidelines for Effective Promotion of Water, Sanitation and Hygiene (WASH) Initiatives in Cambodia.” World Water Policy 5 (2): 118–137.
2. Bartram, J., and S. Cairncross. 2010. “Hygiene, Sanitation, and Water: Forgotten Foundations of Health.” PLoS medicine 7 (11): e1000367. https://doi.org/10.1371/journal.pmed.1000367.
3. Berg, M., C. Stengel, P.T.K. Trang, P.H. Viet, M.L. Sampson, M. Leng, S. Samreth, et al. 2007. “Magnitude of Arsenic Pollution in the Mekong and Red River Deltas—Cambodia and Vietnam.” Science of the Total Environment 372 (2-3): 413–425.
4. Bhowmick, A., K. Irvine, and R. Jindal. 2017. “Mathematical Modeling of Effluent Quality of Cha-Am Municipality Wastewater Treatment Pond System using PCSWMM.” Journal of Water Management Modeling 25: C423. https://doi.org/10.14796/JWMM.C423.
5. Brown, J., and M.D. Sobsey. 2010. “Microbiological Effectiveness of Locally Produced Ceramic Filters for Drinking Water Treatment in Cambodia.” Journal of Water and Health 8 (1): 1–10.
6. Brown, J., and M.D. Sobsey. 2012. “Boiling as Household Water Treatment in Cambodia: A Longitudinal Study of Boiling Practice and Microbiological Effectiveness.” The American Journal of Tropical Medicine and Hygiene 87 (3): 394–398.
7. Brown, J., M.D. Sobsey, and D. Loomis. 2008. “Local Drinking Water Filters Reduce Diarrheal Disease in Cambodia: A Randomized, Controlled Trial of the Ceramic Water Purifier.” The American Journal of Tropical Medicine and Hygiene 79 (3): 394–400.
8. Brown, J., R. Chai, A. Wang, and M.D. Sobsey. 2012. “Microbiological Effectiveness of Mineral Pot Filters in Cambodia.” Environmental Science & Technology 46 (21): 12055–12061.
9. Cassivi, A., E. Tilley, E.O.D. Waygood, and C. Dorea. 2021. “Household Practices in Accessing Drinking Water and Post Collection Contamination: A Seasonal Cohort Study in Malawi.” Water Research 189: 116607. https://doi.org/10.1016/j.watres.2020.116607.
10. Chakravarty, I., A. Bhattacharya, and K. Das. 2017. “Water, Sanitation and Hygiene: The Unfinished Agenda in the World Health Organization South-East Asia Region.” WHO South-East Asia Journal of Public Health 6 (2): 22–26.
11. Clasen, T.F., and A. Bastable. 2003. “Faecal Contamination of Drinking Water during Collection and Household Storage: The Need to Extend Protection to the Point of Use.” Journal of Water and Health 1 (3): 109–115.
12. Clasen, T.F., D.H. Thao, S. Boisson, and O. Shipin. 2008. “Microbiological Effectiveness and Cost of Boiling to Disinfect Drinking Water in Rural Vietnam.” Environmental Science & Technology 42 (12): 4255–4260.
13. Crocker, J., and J. Bartram. 2014. “Comparison and Cost Analysis of Drinking Water Quality Monitoring Requirements versus Practice in Seven Developing Countries.” International Journal of Environmental Research and Public Health 11 (7): 7333–7346.
14. Cui, X., J.W. Talley, G. Liu, and S.L. Larson. 2011. “Effects of Primary Sludge Particulate (PSP) Entrapment on Ultrasonic (20 kHz) Disinfection of Escherichia coli.” Water Research 45 (11): 3300–3308.
15. Daly, S.W., and A.R. Harris. 2022. “Modeling Exposure to Fecal Contamination in Drinking Water due to Multiple Water Source Use.” Environmental Science & Technology 56 (6): 3419–3429.
16. Dany, V., C. Visvanathan, and N.C. Thanh. 2000. “Evaluation of Water Supply Systems in Phnom Penh City: A Review of the Present Status and Future Prospects.” International Journal of Water Resources Development 16 (4): 677–689.
17. Doyle, S.E. 2012. “City of Water: Architecture, Urbanism and the Floods of Phnom Penh.” Nakhara: Journal of Environmental Design and Planning 8: 135–154.
18. Eliyan, C., J.R. McConville, C. Zurbrügg, T. Koottatep, K. Sothea, and B. Vinnerås. 2022. “Generation and Management of Faecal Sludge Quantities and Potential for Resource Recovery in Phnom Penh, Cambodia.” Frontiers in Environmental Science 10–2022. https://doi.org/10.3389/fenvs.2022.869009  
19. Elmore, A.C., G.R. Miller, and B. Parker. 2005. “Water Quality in Lemoa, Guatemala.” Environmental Geology 48: 901–907.
20. Feldman, P.R., J.W. Rosenboom, M. Saray, C. Samnang, P. Navuth, and S. Iddings. 2007. “Assessment of the Chemical Quality of Drinking Water in Cambodia.” Journal of Water and Health 5 (1): 101–116.
21. Higgins, J.A., K.T. Belt, J.S. Karns, J. Russell-Anelli, and D.R. Shelton. 2005. “tir-and stx-positive Escherichia coli in Stream Waters in a Metropolitan Area.” Applied and Environmental Microbiology 71 (5): 2511–2519.
22. Holman, E.J., and J. Brown. 2014. “Safety of Packaged Water Distribution Limited by Household Recontamination in Rural Cambodia.” Journal of Water and Health 12 (2): 343–347.
23. Hunter, P.R., H. Risebro, M. Yen, H. Lefebvre, C. Lo, P. Hartemann, C. Longuet, et al. 2014. “Impact of the Provision of Safe Drinking Water on School Absence Rates in Cambodia: A Quasi-experimental Study.” PloS ONE 9 (3): e91847. https://doi.org/10.1371/journal.pone.0091847.
24. Hutton, G., and C. Chase. 2016. “The Knowledge Base for Achieving the Sustainable Development Goal Targets on Water Supply, Sanitation and Hygiene.” International Journal of Environmental Research and Public Health 13 (6): 536. https://doi.org/10.3390/ijerph13060536.
25. Im, N., K. Kawamura, E. Suwandana, and Y. Sakuno. 2015. “Monitoring Land Use and Land Cover Effects on Water Quality in Cheung Ek Lake using ASTER Images.” American Journal of Environmental Sciences 11 (1): 1–12.
26. Irvine, K., T.P. Murphy, M. Sampson, V. Dany, S. Vermette, and T. Tang. 2006. “An Overview of Water Quality Issues in Cambodia.” Journal of Water Management Modeling R225-02. https://doi.org/10.14796/JWMM.R225-02
27. Irvine, K.N., M. Sampson, T. Visoth, M. Yim, K. Veasna, T. Koottatep, and J. Rupp. 2008. “Spatial Patterns of E. coli and Detergents in the Boeng Cheung Ek Treatment Wetland, Phnom Penh, Cambodia. In: The 6th International Symposium on Southeast Asia Water Environment, Bandung, Indonesia, pp. 78–81.
28. Irvine, K., M.C. Rossi, S. Vermette, J. Bakert, and K. Kleinfelder. 2011. “Illicit Discharge Detection and Elimination: Low Cost Options for Source Identification and Trackdown in Stormwater Systems.” Urban Water Journal 8 (6): 379–395.
29. Irvine, K., S. Vermette, and F.B. Mustafa. 2013. “The “Black Waters” of Malaysia: Tracking Water Quality from the Peat Swamp Forest to the Sea.” Sains Malaysiana 42 (11): 1539–1548.
30. Irvine, K., C. Sovann, S. Suthipong, S. Kok, and E. Chea. 2015. “Application of PCSWMM to Assess Wastewater Treatment and Urban Flooding Scenarios in Phnom Penh, Cambodia: A Tool to Support Eco-city Planning.” Journal of Water Management Modeling 23: C389. https://doi.org/10.14796/JWMM.C389
31. Irvine, K.N., N. Mische, J. Bowles, T. Koottatep, and P. Pichadul. 2016. “Assessing Water Vulnerabilities: Successes, Failures, and Missed Opportunities in a Karen Hill Tribe Village on the Thailand-Myanmar Border.” Journal of Geography, Environment and Earth Science International 5 (3): 1–19.
32. Irvine, K., H.H. Loc, C. Sovann, A. Suwanarit, F. Likitswat, R. Jindal, T. Koottatep, et al. 2021. “Bridging the Form and Function Gap in Urban Green Space Design through Environmental Systems Modeling.” Journal of Water Management Modeling 29: C476. https://doi.org/10.14796/JWMM.C476
33. Japan lnternational Cooperation Agency (JICA). 1999. The Study on Drainage Improvement and Flood Control, Volume 2. Municipality of Phnom Penh.
34. Jensen, P.K., J.H. Ensink, G. Jayasinghe, W. Van Der Hoek, S. Cairncross, and A. Dalsgaard. 2002. “Domestic Transmission Routes of Pathogens: The Problem of In‐house Contamination of Drinking Water during Storage in Developing Countries.” Tropical Medicine & International Health 7 (7): 604–609.
35. Keane, D.A., K.G. McGuigan, P.F. Ibáñez, M.I. Polo-López, J.A. Byrne, P.S. Dunlop, K. O’Shea, et al. 2014. “Solar Photocatalysis for Water Disinfection: Materials and Reactor Design.” Catalysis Science & Technology 4 (5): 1211–1226.
36. Loc, H.H., K.N. Irvine, A. Suwanarit, P. Vallikul, F. Likitswat, A. Sahavacharin, P.C. Sovann, and L.S. Ha. 2020. “Mainstreaming Ecosystem Services as Public Policy in South East Asia, from Theory to Practice.” In: Sustainability and Law (pp. 631–665). Springer, Cham.
37. McGuigan, K.G., R.M. Conroy, H.J. Mosler, M. du Preez, E. Ubomba-Jaswa, and P. Fernandez-Ibanez. 2012. “Solar Water Disinfection (SODIS): A Review from Bench-top to Roof-top.” Journal of Hazardous Materials 235: 29–46.
38. McFadden, M., J. Loconsole, A.J. Schockling, R. Nerenberg, and J.P. Pavissich. 2017. “Comparing Peracetic Acid and Hypochlorite for Disinfection of Combined Sewer Overflows: Effects of Suspended-Solids and pH." Science of the Total Environment 599: 533–539.

39. Mialhe, F., Y. Gunnell, O. Navratil, D. Choi, C. Sovann, J. Lejot, B. Gaudou. et al. 2019. "Spatial growth of Phnom Penh, Cambodia (1973–2015): Patterns, rates, and socio-ecological consequences." Land Use Policy 87, 104061.

40. Mom, K., and S. Ongsomwang. 2016. “Urban Growth Modeling of Phnom Penh, Cambodia using Satellite Imageries and a Logistic Regression Model.” Suranaree Journal of Science & Technology 23 (4): 481–500.
41. Murphy, H.M., M. Sampson, E. McBean, and K. Farahbakhsh. 2009. “Influence of Household Practices on the Performance of Clay Pot Water Filters in Rural Cambodia.” Desalination 248 (1–3): 562–569.
42. Murphy, H.M., E.A. McBean, and K. Farahbakhsh. 2010. “A Critical Evaluation of Two Point-of-use Water Treatment Technologies: Can they Provide Water that Meets WHO Drinking Water Guidelines?” Journal of Water and Health 8 (4): 611–630.
43. Murphy, T., K. Phan, E. Yumvihoze, K. Irvine, K. Wilson, D. Lean, B. Ty, et al. 2018. “Groundwater Irrigation and Arsenic Speciation in Rice in Cambodia.” Journal of Health and Pollution 8 (19): 180911. https://doi.org/10.5696/2156-9614-8.19.180911.  
44. Nakhle, P., O. Ribolzi, L. Boithias, S. Rattanavong, Y. Auda, S. Sayavong, R. Zimmerman, et al. 2021. “Effects of Hydrological Regime and Land Use on In-stream Escherichia coli Concentration in the Mekong Basin, Lao PDR.” Scientific Reports 11 (1): 3460.
45. Nhim, T., and C. Mcloughlin. 2022. “Local Leadership Development and WASH System Strengthening: Insights from Cambodia.” H2Open Journal 5 (3): 469–489.
46. Nowicki, S., Z.R. deLaurent, E.P. de Villiers, G. Githinji, and K.J. Charles. 2021. “The Utility of Escherichia coli as a Contamination Indicator for Rural Drinking Water: Evidence from Whole Genome Sequencing.” PLoS ONE 16 (1): e0245910.
47. Orgill, J., A. Shaheed, J. Brown, and M. Jeuland. 2013. “Water Quality Perceptions and Willingness to Pay for Clean Water in Peri-urban Cambodian Communities.” Journal of Water and Health 11 (3): 489–506.
48. Paretti, N.V., A.L. Coes, C.M. Kephart, and J.P. Mayo. 2018. “Collection Methods and Quality Assessment for Escherichia coli, Water Quality, and Microbial Source Tracking Data within Tumacácori National Historical Park and the Upper Santa Cruz River, Arizona, 2015–16.” U.S. Geological Survey, Scientific Investigations Report 2017–5139, 30 p. https://doi.org/10.3133/sir20175139
49. Percival, T., and P. Waley. 2012. “Articulating Intra-Asian Urbanism: The Production of Satellite Cities in Phnom Penh.” Urban Studies 49 (13): 2873–2888.
50. Phnom Penh Water Supply Authority (PPWSA). 2020. Annual Report 2020. (https://www.ppwsa.com.kh/en/index.php?page=annual-report Accessed 5 Nov. 2021).
51. Poirot, E., S.V. Som, F.T. Wieringa, S. Treglown, J. Berger, and A. Laillou. 2020. “Water Quality for Young Children in Cambodia—High Contamination at Collection and Consumption Level.” Maternal & Child Nutrition 16: e12744. https://doi.org/10.1111/mcn.12744.  
52. Polya, D.A., A.G. Gault, N. Diebe, P. Feldman, J.W. Rosenboom, E. Gilligan, D. Fredericks, et al. 2005. “Arsenic Hazard in Shallow Cambodian Groundwaters.” Mineralogical Magazine 69 (5): 807–823.
53. Prüss-Ustün, A., J. Wolf, J. Bartram, T. Clasen, O. Cumming, M.C. Freeman, B. Gordon, et al. 2019. “Burden of Disease from Inadequate Water, Sanitation and Hygiene for Selected Adverse Health Outcomes: An Updated Analysis with a Focus on Low-and middle-income Countries.” International Journal of Hygiene and Environmental Health 222 (5): 765–777.
54. Ro, C., C. Sovann, D. Bun, C. Yim, T. Bun, S. Yim, and K.N. Irvine. 2020. “The Economic Value of Peri-urban Wetland Ecosystem Services in Phnom Penh, Cambodia.” In: IOP Conference Series: Earth and Environmental Science 561, 1, 012013. IOP Publishing.
55. Rodriguez-Lado, L.R., D. Polya, L. Winkel, M. Berg, and A. Hegan. 2008. “Modelling Arsenic Hazard in Cambodia: A Geostatistical Approach using Ancillary Data.” Applied Geochemistry 23 (11): 3010–3018.
56. Rogers-Brown, J., R. Johnson, D. Smith, and K. Ramsey-White. 2016. “A Pilot Study to Examine the Disparities in Water Quality between Predominantly Haitian Neighborhoods and Dominican Neighborhoods in Two Cities in the Dominican Republic.” International Journal of Environmental Research and Public Health 13 (1): 39.
57. Sahmakum Teang Tnaut (STT). 2021. “Boeung Tamok or Boeung Kobsrov. Facts and Figures #43”. https://teangtnaut.org/wp-content/uploads/2021/04/STT_Boeung-Tamok-report_ENG_Final.pdf Accessed 15 December 2022.
58. Schneider, H. 2011. “The Conflict for Boeng Kak Lake in Phnom Penh, Cambodia.” Pacific News 36 (July/August): 4–10.
59. Sevea. 2017. Access to Drinking Water in Rural Cambodia, Current Situation and Development Potential Analysis. Report (http://www.seveaconsulting.com/wp-content/uploads/2018/01/Sevea-Access-to-drinking-water-in-rural-Cambodia-2017.pdf Accessed 7 Nov. 2021).
60. Shaheed, A., J. Orgill, C. Ratana, M.A. Montgomery, M.A., Jeuland, and J. Brown. 2014. “Water Quality Risks of ‘Improved’ Water Sources: Evidence from Cambodia.” Tropical Medicine & International Health 19 (2): 186–194.
61. Somara, O., and M. Mihara. 2021. “Evaluation of Water Quality and Vegetable Production in Cheung Ek Lake, Cambodia.” Environmental and Rural Development 12 (2): 194–200.
62. Sovann, C., and D.A. Polya. 2014. “Improved Groundwater Geogenic Arsenic Hazard Map for Cambodia.” Environmental Chemistry 11 (5): 595–607.
63. Sovann, C., K.N. Irvine, S. Suthipong, S. Kok, and E. Chea. 2015. “Dynamic Modeling to Assess Natural Wetlands Treatment of Wastewater in Phnom Penh, Cambodia: Towards an Eco-city Planning Tool.” British Journal of Environment and Climate Change 5 (2): 105–115.
64. Stauber, C.E., E.R. Printy, F.A. McCarty, K.R. Liang, and M.D. Sobsey. 2012. “Cluster Randomized Controlled Trial of the Plastic Biosand Water Filter in Cambodia.” Environmental Science & Technology 46 (2): 722–728.
65. Sthiannopkao, S., K.W. Kim, K.H. Cho, K. Wantala, S. Sotham, C. Sokuntheara, and J.H. Kim. 2010. “Arsenic Levels in Human Hair, Kandal Province, Cambodia: The Influences of Groundwater Arsenic, Consumption Period, Age and Gender.” Applied Geochemistry 25 (1): 81–90.
66. Thomas, K., E. McBean, A. Shantz, and H.M. Murphy. 2015. “Comparing the Microbial Risks Associated with Household Drinking Water Supplies used in Peri-urban Communities of Phnom Penh, Cambodia.” Journal of Water and Health 13 (1): 243–258.
67. Thompson, L., J. Vipham, L. Hok, L., and P. Ebner. 2021. “Towards Improving Food Safety in Cambodia: Current Status and Emerging Opportunities.” Global Food Security 31: 100572. https://doi.org/10.1016/j.gfs.2021.100572
68. Trevett, A.F., R.C. Carter, and S.F. Tyrrel. 2004. “Water Quality Deterioration: A Study of Household Drinking Water Quality in Rural Honduras.” International Journal of Environmental Health Research 14 (4): 273–283.
69. Tsujimoto, K., T. Ohta, K. Aida, K. Tamakawa, K., and M.S. Im. 2018. “Diurnal Pattern of Rainfall in Cambodia: Its Regional Characteristics and Local Circulation.” Progress in Earth and Planetary Science 5 (1): 1–18.
70. Vannavong, N., H.J. Overgaard, T. Chareonviriyaphap, N. Dada, R. Rangsin, A. Sibounhom, T.A. Stenström, and R. Seidu. 2018. “Assessing Factors of E. coli Contamination of Household Drinking Water in Suburban and Rural Laos and Thailand.” Water Science and Technology: Water Supply 18 (3): 886–900.
71. Vanny, L., G. Jiwen, and H. Seingheng. 2015. “Phnom Penh’s Municipal Drinking Water Supply: Water Quality Assessment.” Sustainable Water Resources Management 1 (1): 27–39.
72. Visoth, T., M. Yim, S. Vathna, K. Irvine, and T. Koottatep. 2010. “Efficiency of Phnom Penh's Natural Wetlands in Treating Wastewater Discharges.” Asian Journal of Water, Environment and Pollution 7 (3): 39–48.
73. Vuong, T.A., T.T. Nguyen, L.T. Klank, D.C. Phung, and A. Dalsgaard. 2007. “Faecal and Protozoan Parasite Contamination of 13 Water Spinach (Ipomoea aquatica) Cultivated in Urban Wastewater in Phnom Penh, Cambodia.” Tropical Medicine and International Health 12 (2): 73–81.
74. Wang, Y.C., R.C.Y. Ho, C.-C. Feng, J. Namsanor, and P. Sithithaworn. 2015. “An Ecological Study of Bithynia snails, the First Intermediate Host of Opisthorchis viverrini in Northeast Thailand.” Acta Tropica 141 (Pt B): 244–252.
75. WaterAid. 2018. “Rural Water Supply in Cambodia: A Consolidation of Data & Knowledge and Identification of Gaps & Research Needs. Technical Note.” (https://washmatters.wateraid.org/sites/g/files/jkxoof256/files/rural-water-supply-in-cambodia-consolidation-of-data-and-knowledge-gaps.pdf Accessed 7 Nov. 2021).
76. World Bank. 2016. “Kingdom of Cambodia Strengthening Sustainable Water Supply Services through Domestic Private Sector Providers in Cambodia.” Report No: ACS16819. (https://openknowledge.worldbank.org/handle/10986/23769?show=full Accessed 7 Nov. 2021).
77. World Health Organization (WHO). 2021. “Progress on Household Drinking Water, Sanitation and Hygiene 2000-2020: Five Years into the SDGs.” Geneva: World Health. (https://washdata.org/ Accessed 3 Jan. 2023).
78. Wright, J., S. Gundry, and R. Conroy. 2004. “Household Drinking Water in Developing Countries: A Systematic Review of Microbiological Contamination between Source and Point‐of‐use.” Tropical Medicine & International Health 9 (1): 106–117.