Community infrastructure.

Public piped water systems of varying age, quality, and consistency are present in the urban site, Esmeraldas, the intermediate site, Borbón (water plant constructed in 1990 and upgraded in 2006), and the rural-road communities of Timbiré & Selva Alegre (shared and newer system), Maldonado (older system with poorer perceived quality), and Colon Eloy. A public project to install a piped water system in some of the rural-river communities – Colon de Onzole, Santo Domingo, and Zancudo, but not San Francisco – was under construction but not yet operational as of October 2024. Pipeline supply and pressure vary by proximity to plant and elevation, among other physical factors, and so constraints to use can vary both within and between sites [35,36]. Intermittency of access to piped water is another major driver of variation in access and use across the sites, and different daily and hourly water availability has been well documented in the study area [35,37].

Household water access.

Within ECoMiD communities that have public water systems, access to piped systems at the household-level varies. Households in informal settlements (built without government permission, often in flood or landslide prone areas or on otherwise less desirable land that may be difficult to connect to public services), households on the edges of urban areas, and households in rural areas where no piped systems exist are the least likely to have a household connection [35].

Community sanitation systems.

A sewer system with a treatment plant is in place in the urban site of Esmeraldas. The intermediate town of Borbón has sewerage pipes that discharge directly into the river. Pipes have also been installed in parts of some communities, such as Maldonado (rural-road), as a part of housing provided by the “Ministerio de Desarrollo Urbano y Vivienda” (MIDUVI), but the drainage pipes are not connected to any larger system. There are no sewer systems in place in the rural-river communities. Households in communities without sewerage systems for processing black wastewater typically rely on septic tanks, soak pits, or pit latrines, and use large buckets stored in the bathroom and shower areas to flush toilets and bathe.

Coding.

After an initial read-through of the complete transcripts, the coder (MKMP) inductively coded [42] text from the interviews in Atlas.ti [43]. Predetermined structural codes related to the socioecological framework, such as “community WASH” and “household WASH” were also utilized. In vivo codes were used if a phrase captured a key shared expression among interviewees best captured in their own words.

The coder (MKMP) maintained a codebook (S3 Table) with a complete description of the definition for each code. Although only one researcher independently coded the data, the codebook and initial categories and themes were shared early on with project team members, including field team members in Ecuador, as a form of peer debriefing and triangulation. Participants were not asked to review their transcripts to avoid placing additional burden on the mothers. Coding was completed in March 2024.

Analysis.

We followed recommended practices for thematic analysis [42]: analytic memos were used to identify patterns and categories and to relate the codes and themes with the research questions and the socioecological framework [44,45]. Data processing was conducted primarily through metacoding and cutting and sorting [46] facilitated by Atlas.ti software.

Key methods for pattern identification included comparison of codes and categories between 1) mothers living in different communities along the urban-rural gradient and 2) mothers in different strata of socio-economic status, as determined in the initial sampling frame (S1 Table). Comparison was also made between individual, household, community, geographic, and climatic WASH factors. Findings were presented to and discussed with members of the field team and study team for peer debriefing. After themes were determined, the text of the transcripts was revisited to ensure representativeness and accuracy of the themes as a reflection of the data. Freelists were analyzed using a simple count method to tally the frequency of responses [44,47]. Freelist results were analyzed simultaneously with the open-ended questions so that assessment could be made of divergence or correspondence between the data sources.

Climatic scale: Both rainy and dry season conditions affect WASH access.

In the dry season, access to water sources becomes more limited, as rivers become smaller and more distant and rains lessen. Many mothers reported having to rely on less-preferred water sources during the dry season, and having to work harder or pay more to access their usual sources.

“It is difficult to fill up your water [containers]… from here, you have to go and fill the buckets, and you get tired, sometimes it doesn’t rain, and there is no money to buy water, it is hard.” HH 11

“It is difficult when it is the dry season, those who have water, we have to ask them to fill a tank for us, and they charge a dollar fifty for the tank when it doesn’t rain. But when it rains, then even with my [injured] leg I go outside and try to fill my tanks and buckets.” HH 7008

However, in the wet season, when rains are more frequent, heavy rains can lead to flooding and cause the rivers in the region to overflow. Many mothers expressed an unwillingness to install expensive household WASH hardware like cisterns in areas like the rural-river communities, where flooding events are common and likely to contaminate stored water. Mothers from rural-river communities talked the most frequently about flooding in the interviews (Fig 2).

Download:

• PNG
  larger image
• TIFF
  original image

Fig 2. Sankey diagram of select codes by study site.

Study sites were located along an urban-rural gradient, and included Esmeraldas (urban site), Borbón (intermediate site), rural sites accessible by road (rural – road), and rural sites accessible only by river (rural – river). Thicker lines represent greater frequency of code occurrence across all interviews from that site.

https://doi.org/10.1371/journal.pwat.0000368.g002

Geographic scale: Location on the urban-rural gradient has mixed implications for WASH access.

Water storage: Despite the larger extent of public WASH infrastructure available in the more urban locations of the gradient, mothers in the urban site often struggled with water intermittency and expressed a desire for household water storage options like tanks and cisterns to increase reliability of their water supply. Large water tanks and cisterns provide the greatest volume but take up valuable space, while smaller containers like jerry cans and buckets must be refilled often.

Drinking water sources: Even households that have a piped connection to public water do not always report using it as their main source of drinking water. Among the households that participated in the interviews, less than a third reported piped water as their main source of drinking water, while almost half had access to a household piped connection (Table 1).

Bottled water tends to be more expensive when purchased in smaller containers at greater frequency. In the intermediate site, a number of study households were located in informal settlements without access to the public piped system. Many mothers, particularly in the intermediate site and rural-road communities, described the piped water that was available as too dirty to be used for drinking, though piped water was often described as useful for chores (washing floors, clothes, dishes, bathing, cleaning the bathroom, etc.). Mothers in the rural-river communities lacked access to any piped systems, but described easy access to rivers as a benefit.

“Sometimes in the city, you might go without water for three, four days, but here at least from the river we have access to water all the time” HH 7010

Community scale: Infrastructure quality influences maternal WASH preferences and dictates coping behaviors.

The perceived quality of water sources differed across communities within the same geographic area due to infrastructural variability. For instance, the water treatment plants in the rural-road communities of Selva Alegre and Timbiré were newer and tended to have high perceived quality compared to those in Maldonado and Borbón, where the treatment plants were quite old and perceptions of water quality and acceptability [48] were very poor.

“[I would like] the water to come out cleaner, because sometimes it comes out brown, like river water, it comes out dirty.” HH 2125

Although the ECoMiD study does not collect data on user satisfaction with water, adequate water access has been defined by the AAAQ framework in terms of availability, accessibility, acceptability, and quality, where acceptability is defined by color, small, and taste, and quality is determined by health risk [48]. Availability also varies by community [35], particularly in terms of intermittency of access, with implications for which community members need to consider alternatives and backups during outages to piped systems.

Household scale: Reliable, high-quality piped water is the maternal WASH preference at the household level, but WASH source prioritization responds to a number of broader constraints.

Mothers often expressed a desire for access to clean, drinkable water consistently and easily available in their homes for all uses – consumption, cooking, and cleaning. Typically, this access was envisioned as a part of a piped water system.

“[I would like to have] clean water where you can just turn on the hose and see clean water that you don’t have to store… I could use it for the bathroom but also to drink.” HH 3117

In practice, mothers changed their primary water sources based on intended use (drinking vs chores), season, and other conditions, and did not rely on a single source. Mothers described a hierarchy of preferred water sources in response to constraints (Fig 3).

Download:

• PNG
  larger image
• TIFF
  original image

Fig 3. Constraints to maternal water choice, mapped to a socioecological framework.

Constraints shown for the most commonly mentioned water sources or storage containers. Shading matches the socioecological levels indicated in Fig 1: Individual level (purple); household level (yellow), community level (brown), climatic level (green).

https://doi.org/10.1371/journal.pwat.0000368.g003

“When it rains hard there are people who collect their rainwater, and when it doesn’t rain, they use their piped water, and only when that is broken and empty, then they go to the river to collect their water, or they go and ask [someone] to fill their water.” HH 3207

“To get water when it rains, you can collect rainwater, and when you are in a time when it isn’t raining, you have to go get water from the river and boil it, or you have go and buy your bottle of water.” HH 7008

Water for cooking fell into a middle ground in terms of distinguishing between preferred water sources, with some mothers expressing that “dirtier”/less preferred water sources could still be used in cooking, and others noting that they would only use bottled water for cooking.

“In rich households the water is treated and here in the middle class, for example we have to buy one kind of water to be drinking, and they use the same water for everything because their water is better treated, and ours, if we are talking about the tap water, we can’t use it to cook, we have a specific type of water to cook and another kind of water to do chores or clean.” HH 2414

Many mothers described cisterns, elevated tanks, or wells connected to a tubing system and pump in the home as valuable when there was no access to piped systems or when that access was unreliable. In the open-ended interview questions, mothers focused primarily on cisterns as a preferred water storage option, and cisterns were mentioned the most frequently as a WASH priority.

“With a cistern you fill it up, and if there is no water you have your water, you don’t go without.” HH 3207

However, cisterns were considered very expensive and difficult to obtain. In addition, mothers frequently discussed the labor needed to maintain them.

Individual scale: Mothers prioritize time-savings associated with WASH access.

Mothers often discussed prioritizing WASH-related purchases in relation to the time-saving benefit that easier access to water or sanitation in the house provided, particularly as a means to free up time for childcare. For instance, household assets such as washing machines were considered valuable in that clothes washing could take place in the home with children present (Fig 1). Similarly, having onsite access to water was often discussed as being valued because it averted the need to leave the home, with or without young children, to collect water. Conversely, many mothers discussed prioritizing purchasing, boiling, or treating water for their children’s consumption, despite the additional cost, time, and labor required to do so.

We also noted some cross-cutting themes that impacted results at multiple scales.

Perceived constraints to sanitation access centered on water access.

We included several questions on feces management in the qualitative interviews, but mothers were hesitant to address issues related to feces directly. However, there was a general consensus that water is a necessary tool to enable households to maintain a clean and hygienic environment.

“In order to have a clean and tidy house, you have to have water, if you don’t have water, what else can you do except sweep, how can you clean the bathroom?” HH 1046

“Water is essential” [repeated by many HHs] – “Water is the most essential in order to keep the house clean” HH 1149

Most mothers in our study reported owning a toilet that connected to a sewer, septic tank, or pit (Table 1). Few households reported using unimproved sanitation facilities and just two households reported sharing a bathroom with another household. There was a general perception among mothers that household sanitation facilities (primarily septic tanks) were sufficient and without issues – the main issue identified was that water was needed to clean and manage the toilet (e.g., to flush). Mothers did not express concern about their septic tanks filling up or overflowing as long as they could be covered. Some mothers identified sewers as community-level priorities that they would like the government to invest in – but almost always secondarily to playgrounds for children, sports fields, and improvements to roads.

Climatic drivers.

The seasonal impacts of dry and flood conditions dictated maternal WASH choices in different sites in our study (e.g., mothers had to pay for or travel to water during the dry season, and experienced contamination that lessened the appeal of cisterns in flood conditions). In Mexico and elsewhere, year-round water scarce conditions are growing more common, and such scarcity in both access to piped and rainwater has been connected with increased use of WASH coping behaviors [50,51]. Coping mechanisms are generally estimated to be more costly than access to formal infrastructure [52], and so increasing water scarcity is likely to drive increased household WASH expenditures.

Extreme weather events and heavy rainfall are also expected to increase with climate change [53]. Stored water and sanitation systems are both vulnerable to flooding, and while contamination of a stored water source can harm an individual household, contamination from sanitation overflows expands to the community level. Recent research has found that the impact of WASH interventions on health can vary by season [54–56], underscoring the importance of better understanding this driver and providing climate-resilient WASH infrastructure. National level programs that increase climate change resilience, such as those in the Agenda Hábitat Sostenible del Ecuador 2036 [26], can improve the context in which climate-vulnerable communities operate.

Geographic and community drivers.

High levels of variation in quality, acceptability, availability, and accessibility to public infrastructure across and within our urban-rural study sites dictated maternal WASH choices. The piped water access provided to households by community infrastructure was particularly important in increasing the ease of access to large quantities of water. Perceived quality of piped water modified mothers’ water choices differently for consumption and for chores. Water used for hygiene has indirect benefits on health [57], and even contaminated or intermittent piped water can provide important reductions in the labor of water collection. In communities without pipes, purchased water, rainwater, and river water were the only options for water access, and purchased water was considered as important for drinking but not for chores in those communities, likely reflecting poverty as a barrier to expenditures and the lesser importance of quality for domestic tasks.

In some instances, there was a clear discrepancy between items that mothers freelisted as household priorities and the opinions expressed in the open-ended interviews. While bottled water for drinking, rainwater for chores, and cisterns for technology were freelisted as the most important household items, in discussion mothers focused on the limitations of these items and expressed a desire for quality public services. These discrepancies indicate that the WASH options currently available to mothers in our study are not meeting their needs.

Intermittency in piped water systems has been found to erode community trust, decrease water security [58,59], and increase reliance on the use of alternate water sources. A recent study in three of the ECoMiD communities found that intermittency increased reliance on bottled water [35,37]. Prior research also found high rates of bottled water purchasing among poor households in the rural-river communities that lack access to any community infrastructure, in place of home treatment methods [37]. Despite the immense household expense and financial concerns voice by the mothers in our study, increasing reliance on bottled water linked to failed public service provision is a trend across low-resource countries [60]. Water testing conducted by our team and others has found that large reusable bottled water sold in Ecuador is contaminated with coliforms, consistent with other global findings [61]. National investment in both rural and urban water systems is essential to improving community level services [23].

Another infrastructural driver of water choice is access to electricity. When the power is out, household technologies that depend on tube and pump systems, drawing water from wells and cisterns, are not able to distribute water throughout the household. The high costs of providing large-scale electrification programs, particularly in rural areas, present a similar challenge as those for constructing large water distribution and sewage systems. Off-grid solutions, such as decentralized grids or smaller, local solar or hydroelectric projects, have been proposed as community-level alternatives for access [62,63]. However, droughts in the region pose a risk for relying on hydroelectric power. Mothers in our study described storing piped and other water in response to planned or predicted outages to cope with these conditions. In 2024, droughts led to widespread electrical outages across South America, including in Ecuador; with climate change, these events are predicted to increase further [64]. Planned power cuts lasting up to 12 hours a day several days a week were commonly taking place throughout the study period of 2022–2024 [64]. Because many WASH systems, like pumps, depend on electricity, cuts to power also mean cuts to water and/or sanitation services in much of the country. WASH solutions provided by governments or via interventions must be able to overcome these infrastructural challenges. There is increasing interest in household-level solutions that can maintain climate resilience in spite of infrastructural deficiencies. However, our results suggest that resources would be better spent on developing more climate-resilient infrastructure at the community level.

Further, infrastructure and electricity limitations are not isolated to individual communities – political will at the national and international level (e.g., SDGs) to provide such services universally, and national capacity to implement such services, dictate the lived reality at the local level. Although we did not ask mothers about sociopolitical context, mothers often brought up these issues, which are a direct result of national processes. Many water systems in smaller communities, including several in our study, are community-financed, with household payments directly supporting system maintenance. There are inconsistent enforcement mechanisms for payment, and non-payment practices may leave such community systems underfinanced, especially without sufficient support and investment from regional and national authorities [65]. Ecuador has experienced increasing sociopolitical instability and violence since 2020 [66,67], with an influx of international narcotrafficking groups and growing levels of corruption in the government, including in the institutions responsible for providing public services such as WASH infrastructure [68]. The province of Esmeraldas already suffers from a lack of investment and maintenance for public projects [69], which is likely to continue under the current context. Historic exclusion of rural, poor, or racialized communities from formal networks during colonization in the region more broadly, as well as the current exclusion of marginalized groups under capitalism, also dictate community level possibilities for water choices [70].

Household drivers.

In line with prior research on drivers of water coping practices [51,71–77], labor and costs were major drivers of maternal water choice for both consumption and chores in our study. Water costs included purchasing bottled water, paying someone to collect water, paying for monthly piped water access, paying for fuel or materials to treat water, plus the time and labor costs to collect, store, and treat water from rain or rivers. Both financial and labor expenditures were important and consistent themes expressed by our study participants as burdens that they experienced.

Although our study was limited to female heads of household, mothers are not the only residents making decisions about WASH at the household-level. Women are often more limited in their decision making and in their control over household resources compared to the male head of household [78,79]. In our study population, although some participants reported part-time employment, men were more likely to be the formally employed household member. It is likely that women in Ecuador face similar challenges to exercising agency in financial decision making around WASH as those that have been reported in similar settings elsewhere [78,79], and that gender and gender norms are operating at all levels of the socioecological model.

In addition to gender-based competing priorities for household expenditures, there may also be competing needs for the target of water use at the household level, for example, to attend to agriculture or livestock needs. Both these water-intensive activities can provide important income and nutrition for families, while limiting water availability for other household uses. Both practices can also potentially increase exposure to fecal contamination in the absence of good hygiene and animal and/or crop management practices [80,81].

The percent of household income being spent on water for any use may be extensive for ECoMiD households, given the quantity of water most mothers described needing to purchase on a weekly basis, and the low incomes and relative scarcity of employment options in the study communities. It has been estimated that globally, while a high resource country resident may spend less than 1% of their daily income on water, a low resource country resident could spend nearly 50% of that income on water, and private purchasing [60] can be immensely more costly compared to subsidized access [82]. Programs like Ecuador’s cash transfer (BONO) [25] are important for improving poverty levels overall to enable improved WASH decision making.

Small, daily or weekly water purchases can add up to being much more expensive than monthly connections or larger volume solutions like cisterns, something mothers in our study population were acutely aware of [83]. Mothers in our study area tended to purchase smaller quantities of water from neighbors with wells or piped water connections, mobile water vendors, and from stores, in the absence of being able to connect to a community-level piped system or purchase large volume storage. It is difficult to quantify the exact amount expended by ECoMiD participants using existing survey data, but the burden was clearly expressed in our interviews, and prior research in the study area conducted in 2010 reported monthly water bills for piped connections ranging from less than $3 to more than $30, with frequent reports of non-payment [84]. In our interviews, mothers estimated the monthly cost of piped water at $5-$10 a month depending on volume, and households are charged regardless of outages. One mother mentioned that you could be cut off for non-payment, and that getting the connection reinstated was extremely difficult. The 2010 report also found that more poor households owned a piped connection in areas where government subsidies were provided (Timbiré) compared to in areas without them (Borbón), again highlighting the role of government in community access [84].

Household technology has also been put forward as an additional WASH choice and determinant of subsequent choices [73,85]. In the water realm, technology includes hardware such as cisterns, toilets, water tanks, pumps, and wells that households can purchase or construct. We identified hesitancies in mothers to invest in expensive WASH technologies particularly in houses they did not own, and in informal settlements. Unwillingness to invest in houses in informal settlements is a barrier that was identified as a major constraint to wealth generation in Latin America by Hernando de Soto [86], who promoted legalization of informal settlements as the path towards economic growth and poverty alleviation. In the absence of formal home ownership, local WASH non-governmental organizations in the area have suggested that affordable, temporary alternatives, such as large, transportable plastic tanks, could be a potential solution, but more research with individuals and communities is needed to ascertain if households would want to invest in something more immediately accessible but possibly of perceived lower quality/durability compared to long-term hardware solutions (Personal Comms, Green Empowerment).

Individual drivers.

Our findings that mothers make tradeoffs in their WASH decision making, including decisions about utilization, purchasing, and investment, is consistent with existing literature [51,77]. In the context of water choice in low-resource, high-contamination settings, risk of infection is a key part of these tradeoffs [87]. In many instances in the interviews, mothers expressed an awareness that they were making these choices. The frequent discussion of purchasing, treating, or boiling water for young children indicates there is maternal willingness to expend extra time and/or money to achieve perceived health benefits, particularly for this vulnerable age group, but these investments may not fully protect children from infection in this context.

Mothers in our interviews also used social capital to overcome constraints to water access [88], by borrowing water or toilet access from friends, family, and neighbors. Water and toilet borrowing are common coping mechanisms that do not require financial expenditures [88], but which have been connected to increased stress [77]. Water was described by mothers as essential for hygiene in the open-ended interviews, but freelisted less frequently than chlorine and disinfectant. In this case, it is possible that water was viewed as so fundamental a need that it did not always occur to mothers to list it as an “item” they might need, or perhaps the focus on items that decontaminate underscores the connection mothers make between hygiene and health. Either could explain the discrepancy in the findings observed between the methods.

Women tend to be responsible for making WASH-related decisions at the household level [77], while men are more likely to lead infrastructure projects and administer urban water and sanitation systems, ultimately dictating the broader structures of access [89]. This gender disparity has important implications for the limitations placed on women in the WASH process, where they are so often asked to make tradeoffs between costs, labor, and health [77], and their relative inability to alter the broader structural factors at play. If individual preference is rarely considered in designing WASH interventions [17], maternal power and needs may be even less so [7,13,79]. Current efforts need to do more to center the women who bear the brunt of the burden for accessing WASH. A recent review of women’s engagement in WASH interventions found that all interventions included were either gender unequal or unaware [13], meaning that they ultimately did not address the burden on women in providing WASH access for their families. Centering the financial and labor costs of women is likely to lead to more effective WASH interventions. Understanding the various limitations placed on women’s agency at every level of the socioecological model could help direct efforts at empowerment to specific and context-relevant areas of women’s lives.

Intersectional drivers.

WASH choices are further dictated by broader contexts that impact each level of the model: Ecuadorian historic, socio-political, economic, and cultural factors, including racialization and gender dynamics, dictate the WASH choices and finances available to communities, decision-making parameters within households, and ultimately the agency of the individuals that live within them. Together, our results and the application of the socioecological analysis suggest that continued and increased public investment in climate resilient infrastructure for rural and urban WASH, poverty alleviation programs focusing on female empowerment, and housing and other multisectoral approaches to health and wellbeing are needed to support individual WASH-related decision making and to improve related health behaviors and outcomes.

Money is a clear enabler of increased options for WASH access and presented households with the means to overcome infrastructural failures and seasonal calamities, among other challenges. Ultimately, household wealth is a downstream outcome of numerous higher-level influences, from historic colonization practices to government investment decisions and national economic resources. Control over a household’s financial resources is also dictated by social and gender norms, as described earlier in this paper.

Lack of access to WASH burdens the poorest households the most – poor households are already less likely to have access to public WASH connections [15], and the poorer quality that access is, the more likely that a household will need to expend additional money or labor to access alternate water sources. The burden of obtaining WASH access globally is highest on the poorest [50,76], and coping costs, financial and labor-related, place the largest burden on the poorest populations and on women [76,90]. Weekly purchasing of water requires both time and money, and increases unpredictability [50] and stress [91], as reflected in our interviews, with likely biological consequences associated with heightened stress, particularly for women.

Installation costs for cisterns, tanks, and other hardware that could maintain access in the face of intermittency or low-quality piped water were high and perceived as insurmountable by many of our study participants. Seasonal droughts and flooding compounded the financial burden on households to purchase or expend extra labor to acquire water, particularly in the rural-river communities without infrastructure. Yet the burden placed on the poorest exists across the urban-rural gradient, due to the variation in infrastructure quality and consistency described above. Infrastructure that is accessible, reliable, and high quality can help reduce the burden on households across wealth levels.

Many mothers suggested that there were limited safe drinking water choices available to even the wealthiest residents in our study sites. As such, it is possible that the majority of the ECoMiD households face similar levels of exposure to unsafe drinking water, but the poor are expending more of their resources to access that water. These expenditures may leave the poorest households more vulnerable to other interconnected challenges of poverty, such as food insecurity and unequal access to electricity, and increase the cumulative risk of poor health outcomes, such as growth shortfalls and stunting. Though the health benefits of improved WASH are clear [57,92–95], our research shows that the choices individuals make in utilizing and investing in WASH are dictated by higher-level factors, including ability to access public infrastructure or to pay for alternate means of accessing water, consistent with other study findings [7,96], and underscores the limitations of assessing WASH access only through a single metric (e.g., a piped connection in the household).