Community finance and promotion of Managed Aquifer Recharge systems affect uptake and sustainability as a potable water source in southwest coastal Bangladesh: A qualitative study

In the south-western coastal region of Bangladesh ground water, normally used for drinking, is excessively saline. Increased salinity in drinking water can have a range of adverse health impacts. Managed aquifer recharge (MAR) systems, which infiltrate rainwater and fresh pond water into aquifers, are adaptive strategies to deliver low-saline water to the affected communities. Several MAR systems have been piloted in these regions. The MAR pilot study demonstrated the potential for increasing freshwater availability and sustainable year-round drinking water supplies. The objective of this study was to identify MAR system management shortcomings and strengths to provide recommendations that are applicable for future MAR systems and engineering driven water management methods in low-income countries. A We


Introduction
The south-west coastal Bangladesh faces a crisis of fresh drinking water due to ground water salinity [1,2] This region is more vulnerable to safe drinking water shortage than elsewhere in the country because both surface and ground water suffer from acute and high salinity intrusion [1]. Pond water, which is microbiologically contaminated [3], is a common low-saline alternative drinking water source in this region [4,5]. Ponds can also be saline, especially after inundation from severe storms [6]. Groundwater salinity has been associated with increased sodium intake through drinking water in coastal Bangladesh estimates suggest up to 16 gm sodium intake per day only through drinking water [7]. Increased salinity in drinking water is likely to have a range of adverse health impacts [8][9][10][11][12]. Epidemiological studies have demonstrated that high sodium intake is associated with elevated blood pressure and cardio vascular diseases [13]. Moreover, drinking pond water that is contaminated and unclean poses a risk for diarrheal and other water-borne diseases. Filtration methods such as biosand or pond sand filters (BSF/PSF), a point of use (POU) water treatment method, can be used at individual household or at community level [14,15]. To meet fresh drinking water crisis needs, different strategies were taken in the coastal areas of Bangladesh including PSF [16]. Community engagement is important when selecting new technologies to encourage uptake and to sustain their use [6,17].
To mitigate drinking water salinity, an adaptive strategy known as Managed Aquifer Recharge (MAR), has been demonstrated in a series of successful installations across south-west coastal Bangladesh. Freshwater derived from monsoon rain is injected into shallow coastal aquifers [18]. The Geology Department of the University of Dhaka, in collaboration with Unicef piloted 18 MAR projects, in three districts of south-west Bangladesh; Khulna, Shatkhira and Bagerhat [ 6].The MAR pilot study demonstrated the potential for increasing freshwater availability and sustainable year-round drinking water supplies. Subsequently 80 MAR systems have been installed in these regions and a health impact evaluation of MAR water access on blood pressure and kidney function has recently concluded [1, 6,19,20].
In this study we enrolled members from 18 MAR pilot communities to explore water source preference, MAR water and system acceptability to explore factors affecting MAR use. The objective of this study was to identify MAR system management shortcomings and strengths to provide recommendations for future management that are potentially applicable to other engineering driven water management methods in low-income countries. Study design and sampling: We used a qualitative research approach, consisting of indepth interviews (IDIs), key informant interviews (KIIs) and focus group discussions (FGDs). To maximize data variability, the sampling framework included three MAR sites per district selected based on groundwater salinity, sites categorized as having high, low and median electrical conductivity (EC) at pilot phase, giving a total 9 pilot sites (Table 1).

Methods
The Focus group discussions included the technical supervisors and caretakers (recruited from same community) of all 18 pilot MAR sites, for a total of 3 FGDs; one in each of the three districts. Our key informants included the key personnel from the Dhaka University team who are responsible to oversee the MAR implementation in the field. IDI interview and FGD sample sizes were determined from data saturation attainment in previous water, sanitation and hygiene qualitative studies [21][22][23].  Data collection : Interview guidelines were developed according to study objectives: identifying MAR system management shortcomings and strengths, to provide recommendations. This was done for each data collection method guided by the Integrated Behavioral Model for Water, Sanitation and Hygiene (IBM-WASH) theoretical framework.
IBM-WASH integrates the multi-level factors influencing water, sanitation and hygiene behavior, in three dimensions; contextual, psychosocial and technological [24].
Four researchers with experience conducting qualitative research carried out data collection in Bengali, the local language. During data collection, we explore their perspective on MAR water including their views on taste/quality, barriers to using and maintaining MAR system, their opinion on sustainability and maintenance of MAR sites as well as sources of drinking/cooking water during the dry season, and the cost of managing fresh drinking water in dry season. Through IDIs with the former users and never-users, we also sought their opinion on the MAR system, reason for non-use and suggestions to improve MAR systems to increase the number of MAR users. The three focus group discussions with the Technical Supervisors and caretakers from all of the 18 pilot MAR sites aimed to assess their understanding about the MAR system, their perspective on the community's attitude towards the MAR system, about the intervention's strengths and weaknesses to elicit suggestions for sustainability and strengthening the intervention.

Data Analysis
All the interviews and group discussions were recorded using digital recorders after taking written consent from the participants prior to commencing. We manually identified code themes according to our research objective and summarized relevant information in English directly under each code by listening to data on the recordings. Each of the interviews and discussions were separately summarized following the same method. We compiled all data under each theme for all the IDIs, FGDs and KIIs separately. We conducted thematic content analysis and provide a brief description of our findings.
We analyzed these data using the IBM-WASH framework dimensions and levels. Where applicable, we analyzed the results according to the contextual, psychosocial and technological dimensions, at the community, behavioral, interpersonal, and individual levels.

Ethical considerations
Data collectors explained the research study objectives clearly to study participants reading the content of consent forms. Before taking part in the study, these participants provided written informed consent which was documented by a signature. The protocol for this study was reviewed and approved by the icddr,b Institutional Review Board, protocol number PR-15096.

Results
Respondents' socio-demographic characteristics IDI respondents were mostly female (19/24), one third were between 18-28 years old and another third was between 28 to 38 years of age. Almost half of the respondents completed primary level education. Two-third of our respondents were homemakers and one third of the households earned US$ 62.50 to US$ 125 per month. One-third of the households were living in the community for between 21 and 40 years ( Table 2). Among the 28 focus group discussion participants (technical supervisors and caretakers recruited from same community) more than half (16/28) completed a bachelor's degree and more than half (16/28) were living in the same community where they were managing MAR site for more than 25 years. The both key informants (MAR installation and management staff) had post graduate degrees.
As we found quite similar responses between the groups; current user, former user and never user, we present combined findings.

Salinity impact on human and animal health
An important motivator for developing the MAR system and therefore an important element of MAR management was the perceived health benefits of drinking low saline water. Thus, we explored health impact perceptions. More than half (13) of respondents reported that drinking saline water causes diseases such as diarrhea, dysentery, gastritis, digestive or abdominal disorders; none described impact on blood pressure. Some (4) could not name a specific health hazard related to water salinity. About one third of respondents (9) mentioned that saline water also affects animal health and has a negative impact on the environment. They said that livestock (cows and goats) suffer from diarrheal disease from drinking saline water. Two respondents mentioned that salinity hampers optimal crop and plant growth. One mentioned that fish die in saline water. However, one current user from Khulna mentioned that, "Saline water is not so harmful because saline can be removed from the body but it is not possible to remove arsenic" Preferred water sources for drinking More than two thirds (13/18) of current MAR users reported MAR as their preferred drinking water source, 3/18 users and 1/3 former user preferred rain water ( A range of problems were described in collecting water, and some respondents found collecting water a positive experience. Some of the users (4) were not paying for MAR use but they mentioned some of their demands and they told that they will be willing to pay in the future if those demands are fulfilled. Some of their demands were; if water quality improves and all the villagers agree to pay monthly USD 0.13 to USD 0.25, and if MAR water remains available the whole day; then they will be willing to pay for MAR water in future. One respondent stated, "We are not paying willingly. Previously we did not pay any money. If MAR provides water for the whole day then we will pay (USD 0.25)20 taka willingly." In contrast, key informants (who oversee the MAR implementation) and focus group participants (MAR management supervisor/caretaker) thought that management and maintenance issues were their primary concern, which were impeded by limited funds, as users do not pay regularly. He said, "Irregular payment of money creates problem in maintenance and management of MAR technology."

Discussion
Communities in south-west coastal Bangladesh are vulnerable to the health impact of saline water consumption and cannot rely on ground water, ponds, or ad-hoc rainwater collection to meet their drinking water needs without involving technologies. Communities' drinking water practices are explained by the contextual dimension at the community and individual level of the IBM-WASH framework ( Table 3). -Initial introductory talks provided by MAR installation staff; some MAR systems have signs in place to describe the system, its use and maintenance.
-MAR systems infiltrate 'fresh water'** into groundwater aquifers to provide low salinity drinking water -MAR systems were installed with funding from Unicef. Households use their own collection vessels.
-The recurrent cost for MAR water usage to be obtained from households.
-Care-taker maintains and operates the system on a daily basis and the technical supervisor visits the site once a week, continuously monitor the rate of infiltration and abstraction -Care-taker records the amount of water infiltration and abstraction every day, tests the MAR water turbidity daily and cleans the filter chamber once a month Community -Pond water is low saline during wet season, used for drinking and cooking. Pond water is contaminated but some households treat this water -Rainwater used for drinking in wet season.
-Rainwater collection has been encouraged as a salinity mitigation method -Shallow tubewell water perceived as low salinity, used for drinking -MAR water is encouraged for use in dry season as tubewell water is high saline and pond water can be high saline.
-User committee formed; functions include collecting maintenance costs, appointing caretaker and technical supervisor, responsible for community mobilization and financial assurance -Neighbors describe the MAR system and its use to those who did not attend introductory meetings -Sign boards at some MAR systems provide information on how the system works -Project funds maintenance until MAR water meets standards for salinity, arsenic, iron and fecal coliforms. The site is then handed over to the User Committee -Household members draw water from MAR systems using a hand-pump, which is a familiar technology. Sometimes there are queues.
-Community members that live more distant from MAR systems are less likely to use them.
-MAR maintenance supervisor responsible to operate the system to recharge the aquifers to provide low saline drinking water in the dry season.
Interpersonal/ Household members and neighbors -Users thought that MAR water quality is better than other drinking water sources.
-Though having limited -No active promotion of MAR water after the initial introductory meetings. No measures to enhance uptake such as a 'refresher' -Some complained that the MAR system was locked at certain times.
-Some complained that water quality was poor knowledge of the impact of drinking saline water on health, but perceived few visible effects related to drinking saline water.
-Female is mainly responsible to manage water for household use, including drinking water.
meeting to capture those who missed the first -Some actively avoid the peak water collection time and the queues Individual -Some households in the MAR catchment area are low socioeconomic status [About two-third of the households (among study participants) earned up-to 10,000 taka per month; selected to represent these income ranges.] -Information on water quality not provided to the community -Pond water was considered low saline and was easy to obtain, close to households.
-The recurrent cost for MAR water usage to be obtained from households.
-MAR system caretaker is unpaid -MAR system caretaker/supervisor collect funds from households for MAR maintenance and operating costs. .

Behavioral/ Habitual
-During dry season, households typically used pond and shallow tubewell water for drinking and cooking before the MAR system was installed.
-Some respondents reported that they used a method to treat pond water -Household members perceived that treated pond water was safe.
-The MAR system was sometimes located at some distance from the household -Respondents reported concurrently using different water sources for different purposes and different sources by season *saline water = electrical conductivity (EC) > 700uS/cm, **freshwater = EC < 700uS/cm The MAR systems were still functioning and being used by adjacent communities 6 years after installation. However, monthly funds that community members reported paying were extremely low and sometimes zero; insufficient to cover operation and maintenance costs. Funding is a major barrier to system sustainability and this concern was mentioned by MAR supervisory staff, with implications for sustained caretaker retainment.
MAR management related findings are explained by the contextual and psychosocial dimension at community/interpersonal level, and technological dimension at the community/individual level in the IBM-WASH framework ( are not accustomed to paying for drinking water [26,27]. Rainwater is considered readily available and to have no cost [28], but water storage vessels need to be provided by households thereby limiting collection capacity especially among the poor [29,30]. Low water tariffs have been described as a barrier to efficient water management in low income countries [31]. Further research is needed to explore mechanisms for equitably funding recurrent MAR system costs and potentially that of a permanent caretaker to enhance maintenance and in turn sustainability; this applies to any technology that is planned for salinity mitigation in Southwest coastal Bangladesh and low-middle income countries with similar interventions that include water and hardware management issues.
Without a more robust funding mechanism, sustainability in the absence of project support is unlikely. Economic studies to determine funding mechanisms are warranted.
In this study, we found that none of the community members described a link between salinity and blood pressure, heart diseases, other circulatory diseases or adverse pregnancy outcomes.
Increasing community health impact awareness has the potential of increasing the number of paying users. A further barrier was limited knowledge; only one third reported that they knew how the system works, and most heard about MAR from their neighbors or others from their locality/village, but not from a social behavioral change communication (SBCC) campaign. Encouraging MAR water uptake could benefit from SBCC and more compelling, potentially marketing-based promotional strategies. Our data suggest that there is a need to better inform the community about advantages of the system, the impact of salinity on human and animal. Messages on how the system works may decrease concerns about water quality, particularly allaying fears about using water from ponds that are contaminated with human and animal excreta. Promotion should include informing the community that the system is resistant to contamination during storms (runoff) and tidal surges (runoff plus salinity), the volume of water that households can collect each day to allay fears of running short. Water taste was found important, similarly elsewhere in Bangladesh [32], thus acknowledging low saline water taste in addition to health benefits could enhance uptake. More compelling SBCC delivery could include displaying messages at MAR sites, rotating the messages to maximize impact or promoting water collection as a social occasion [32].
Community members stated different barriers related to MAR water use/collection, including limited timing to collect water and location of MAR site. Convenience is important for water technology acceptance, as found among other communities in Bangladesh [32]. Thus, solving issues related to inconvenience or providing compelling reasons for the inconvenience are important to consider.
Financial incentives may help. Lower acceptability of water from a shared source has been described [32], however individual user systems can also fail. When a concrete bio-sand filter was provided to individual households in the Dominican Republic, 10% of households were no longer using filtered water one year after installation [33].
Water systems that require continuous operation and maintenance are rarely sustained by local communities in the post-project period when logistical and/or financial support end [33][34][35], and when technical issues are encountered [36]. Strengthening management and maintenance of MAR system could optimize its water use. Methods to enhance community maintenance of water systems after project staff exit can be applied elsewhere.
Limitations to the study include collection of data among a small number of participants. However, to maximize variability in participants and hence their responses, we selected sites from across the three districts where MAR systems were implemented and included sites across aquifer salinity levels. The field team enrolled different types of households based on distance from the MAR system and user category. We also tried to ensure variation across household economic status. Because our study participants had similar practices, along with similar perceptions about the MAR system and water, our findings from this formative study provide a foundation to enhance management systems, which could be applicable for similar settings and other engineering-based water treatment and distribution technologies.
In summary, the concerns about management related to finance, in addition to the amount collected are the greatest threat to system functionality and sustainability. There seemed to be a wide range of payment methods and costs; consistency and equity are important. Additionally, many were concerned about the way in which funds were being managed. It is critical that a committee has guidelines on how to spend the collected sums and to convey transparency to the community members. Developing a simple governance guide for committees to follow would be ideal.
Management/user committee can encourage technology uptake, monitor quality, keep the system operational and recover costs.

Ethics approval and consent to participate
Data collectors explained the research study objectives clearly to study participants reading the content of consent forms. Before taking part in the study, these participants provided written informed consent which was documented by a signature. The protocol for this study was reviewed and approved by the icddr,b Institutional Review Board, protocol number PR-15096.

Consent for publication:
The authors gave their consent for publication of this original research work.

Availability of data and material:
The authors ensure the availability of data and material of this research work and readers can access the data upon request.
Funding: This research study was funded by Wellcome Trust, donors grant # 106871/Z/15/Z.

Authors' contributions
FAN developed the related qualitative part of the research protocol and all authors provided input in its design and methodological development. FAN developed the data collection instruments. LU provided conceptual guidance for data analysis and manuscript development. AS and TJ collects/supervises the data collection, analyzed and summarized the data. FAN drafted the initial manuscript and all authors commented on the drafts and approved the final manuscript. MAN and MR conceptualized the study and secured the funding. Acknowledgements: