Rainwater harvesting (RWH) is increasingly utilized today by populations to alleviate water supply issues, particularly in rural, dry environments. Limited research has considered, simultaneously, the numerous factors that contribute to sustainability – for example, social acceptance, water quality, and maintenance needs – of RWH. This research aimed to improve the understanding of factors influencing the sustainability of rainwater harvesting systems for domestic use (DRWHS) through examination of social, water quality, and technical feasibility components. We conducted 50 household surveys and 17 rainwater quality analyses in San Jose Xacxamayo, a rural, semi-arid community in Puebla, Mexico. Results showed that DRWHS are socially accepted primarily because of the presence of existing local skills and knowledge, as well as critical need for water. Results from most of the water quality parameters measured were within World Health Organization (WHO) guidelines for human consumption, with the exception of pH, total coliform, and heterotrophic plate count, which required water treatment prior to consumption. Technical feasibility was the main barrier to the sustainability of DRWHS; highly seasonal rainfall and small roof sizes (averaging 70 m2) resulted in households unable to meet annual water needs. Increasing roof sizes and providing water treatment could ensure DRWHS sustainability in the studied community.

## INTRODUCTION

Fresh water is essential for life and human well-being, however, populations in many regions worldwide lack basic access to a safe and constant supply. The United Nations estimates that, globally, nearly one billion people lack access to an improved source of drinking water (UN 2012). Individually, several countries are facing conditions of water stress, where annual water availability per capita ranges from 1,000 to 1,700 m3 (Brown & Matlock 2011). Cumulatively, these countries’ water stress conditions currently affect roughly one-fifth of the global population (Arnell 2004; UN 2007; Neibaur 2015). There is a global decline of water supply and water quality in large part because of a growing population (expected to reach 8 billion in 2025), water pollution from human activities, and effects of climate change (Alcamo et al. 2000; Imteaz et al. 2012).

In much of the developing world, populations in rural, semi-arid and arid regions often are more vulnerable to the challenges of access to water because they reside in areas of little rainfall and may face economic, political, or technological barriers to pipe and distribute surface or groundwater (UN 2007; Neibaur 2015). Because of current and future water challenges, there is a need for alternative water supplies that can sustain populations in rural semi-arid or arid areas, enable economic development, and conserve existing surface and groundwater resources (Fuentes-Galván et al. 2015; Lizárraga-Mendiola et al. 2015).

Rainwater harvesting (RWH), an ancient technology used to capture and store rain, dates back over 4,000 years. Evidence shows that RWH originated in the Middle East and Asia, and several countries – notably India, Japan, China, and Turkey – have adopted this practice for over 2,000 years (Gould & Nissen-Peterson 1999; Neibaur 2015). Historically, RWH has been utilized to respond to both agricultural and domestic water supply needs. RWH is currently being revived and utilized in several dry, poor, and rural regions of the world for domestic water provision, as an alternative technology to satisfy populations’ present and future water demands under climate change uncertainties (Gould & Nissen-Peterson 1999; Meera & Ahammed 2006; Barron 2009; Mun & Han 2012). Several international development agencies, such as the United States Agency for International Development (USAID) and the US Peace Corps, support RWH system implementation by working directly with the communities to help facilitate social acceptability of RWH. Furthermore, RWH has been applied as an adaptive measure to address climate change impacts on precipitation variability in places where human populations are heavily dependent on rain-fed local water resources (Ndiritu et al. 2011). Numerous RWH studies have demonstrated improved livelihoods and increased water supply through the implementation and use of DRWHS (Mutekwa & Kusangaya 2007; Barron 2009). Additionally, the utilization of DRWHS, which operate at a household level, is currently being acknowledged as a contemporary, sustainable solution to meet household domestic water demands (Gould & Nissen-Peterson 1999; Mun & Han 2012; Lizárraga-Mendiola et al. 2015).

Numerous factors affect the sustainability of DRWHS. These factors can include social acceptability, water quality, and ability to meet household demands given spatial and temporal variability of rainfall (Neibaur 2015). Social factors, for example, can strongly influence a rural population's willingness to adopt DRWHS, even where there is high need for improved water supply (Fuentes-Galván et al. 2015). Social factors include local beliefs and attitudes, customs, gender roles, economic benefits or costs, local knowledge, materials and technical capacity, among others (Fuentes-Galván et al. 2015). Acceptance of rainwater for potable use can also be dependent upon the user's perception of taste, odor, and appearance (Doria 2010). Further, local customs can impact the reliance on senses to select a water source that can lead to the continued use of more contaminated water, foregoing the opportunity to change behavior and opt for a more reasonable and improved supply option, possibly through RWH (WHO 2004; Neibaur 2015).

Water quality is an additional important factor that affects feasibility of DRWHS, as it directly relates to human health (Baguma et al. 2010). Much literature shows that rainwater alone is typically of high quality; once rainwater interacts with the atmosphere, roof, or storage tank, contamination can occur (Choudhury & Vasudevan 2003; WHO 2008; Helmreich & Horn 2009; Schets et al. 2010). Previous studies generally have concluded that the physiochemical properties of rainwater are within World Health Organization (WHO) guidelines with the exception of pH (Pushpangadan & Sivanandan 2001; Chang et al. 2004; Neibaur 2015). Heavy metal contamination of rainwater is highly dependent on the roof and tank material (Chang et al. 2004; Ward et al. 2010). Zinc, copper, and metallic roofs are reported to negatively affect water quality (as considered for human potability), while water collected from concrete, tile, and aluminum roofs typically is acceptable for potable use (Ward et al. 2010; Mendez et al. 2011). Several other studies of RWH have reported that stored rainwater is contaminated with bacteria that exceed WHO guidelines, a problem which can be attributed to poor system design and infrequent maintenance (Ward et al. 2010; Domènech et al. 2012).

Technical feasibility and reliability of DRWHS refers to the capacity of these systems to meet household water demands. Factors that influence DRWHS efficiency include temporal distribution of rainfall, user demand, catchment size, and storage capacity (Lizárraga-Mendiola et al. 2015). Several studies have examined various optimization methods for DRWHS on the capacity to meet household water demands (Youn et al. 2012; Wallace & Bailey 2015). The behavioral method, developed by Jenkins et al. (1978) and later adapted by Fewkes & Butler (2000), is used to compute volume yield of a tank in a given time, which is a function of rainfall, catchment size, runoff losses, tank size and user demand. The behavioral method can provide accurate and continuous results for the DRWHS input, output and volume in a defined period of time (Fewkes & Butler 2000; Liaw & Tsai 2004; Imteaz et al. 2012). The applied simulation method (Imteaz et al. 2012) aims to determine the rainwater supply delivered by the tank in meeting the users demand over a given time. Reliability determines the frequency that the demand is met and can be examined on a time-based scale (Palla et al. 2011).

Climate change is projected to affect future precipitation patterns and, therefore, the feasibility of DRWHS. In 2025, it is estimated that 2.9 to 3.3 billion people worldwide will be affected by water stress (Arnell 2004). Understanding the long-term feasibility of DRWHS as an alternative to meeting water supply needs for human populations requires evaluation of future precipitation trends and projection of the ability to meet a household's year-round demands (Lizárraga-Mendiola et al. 2015). It also requires an understanding of criteria for social acceptance of DRWHS (Fuentes-Galván et al. 2015).

Numerous studies have examined individual components of DRWHS and their feasibility, however, few studies have investigated social, environmental, technological, and climate factors simultaneously (Youn et al. 2012; Wallace & Bailey 2015). Therefore, considering potential interactions between these factors, current understanding of the reliability of DRWHS for human populations is limited. In the present study, we aimed to address these limitations by examining the sustainability of the DRWHS in a semi-arid, rural community in Puebla, Mexico, through an analysis of social, water quality, and technical feasibilities of the systems. Additionally, we considered how future precipitation patterns could affect the ability of DRWHS to deliver adequate water supply.

## METHODS

We conducted this study in San Jose Xacxamayo, which is a small community located in the municipality of Puebla, the fourth most populous city in Mexico. The community had a total of 827 residents in 2010 (INEGI 2010; OECD 2013). Average annual precipitation is 724 mm, with most falling during the rainy season between May and October (CONAGUA 2012). Our methods for the present study included: (i) surveys with DRWHS users, (ii) water quality sampling of stored rainwater, and (iii) water balance modeling to determine current and future reliability of DRWHS under future climate conditions.

This project was conducted as a part of Peace Corps Volunteer work by E. Neibaur in San Jose Xacxamayo (2012–14) that involved the installation of 82 DRWHS of 10,000 L capacity in 70 different households. Of those 70 households, 50 were randomly selected for in-person surveys. All survey questions were developed and conducted in line with the study's objectives, relevant literature, and approved by the International Review Board (IRB) of Florida International University. Surveys were conducted through personal interviews with questions related to: socio-economic characteristics of households, current water supply sources, economic costs of water, household water demand, understanding of RWH, and the durability and required maintenance of existing DRWHS. Survey questions aimed to understand social acceptability factors, behaviors, attitudes, and skill level of household users with respect to DRWHS. In addition, surveys were used to calculate the DRWHS current and future potential storage capacity and reliability as determined by the roof size, current domestic water demand, and monthly rainfall input. Descriptive statistical tests of survey data were conducted using the Statistical Package for Social Science (SPSS).

Water quality analyses examined whether stored rainwater met conditions for potable human use. From August to October 2014, during the final phase of E. Neibaur's Peace Corps Volunteer assignment, 17 rainwater samples were collected from select DRWHS. Households surveyed confirmed that DRWHS sampled had been cleaned out prior to the rainy season and were used only for storage of pure rainwater (no cross source contamination). Physiochemical and bacteriological analyses were performed on water from the 17 DRWHS. Most water samples were tested and collected directly from stored water in the DRWHS; however, there were a few cases where the water levels were low and a bucket (the usual methods of extraction by households) was utilized. All samples taken were stored in sterilized polyethylene plastic containers. Some water quality parameters were analyzed in situ (temperature, pH, total dissolved solids (TDS), electric conductivity (EC) and other water quality samples (ammonium, phosphorus, heterotrophic plate count (HPC), and total coliform count) were stored in a cooler at 4°C and transported the same day for analysis at a laboratory facility, the Meritorious Autonomous University of Puebla (BUAP). Water quality data were compared against the WHO's standards for water use and human consumption. For absence of criteria standards from WHO, information from previous studies of rainwater quality was compared against our data (Domènech et al. 2012; Adler et al. 2013).

As part of the survey, domestic water demand for weekly use was calculated as reportedly used for: drinking, cooking, washing clothes, bathing, washing dishes, washing the house, and watering plants. Initially, our question considered all domestic uses for the household per week. For the initial eight surveys conducted, responses were conservative and with high uncertainties. Thus, the additional 42 surveys recorded the weekly domestic water demand separately for each domestic use per household, allowing a more precise response by the participant. The total domestic water use for each household was used to calculate the water volume required to be captured and stored given the monthly precipitation in order to meet the family's need. In addition, monthly demands and roof sizes of all surveyed users were averaged to calculate the sample population's overall average feasibility and reliability of RWH. Rainfall data were extracted from the nearest meteorological station, Balcon Diablo A. Texaluca (18°89′75″N−98°13′22″W) from 1951 to 2010, a 49 year time period (INEGI 2010).

Calculating the volume of potential rainwater collection is important to optimize the size of the rainwater tank for sustainable water use. The key component to a DRWHS is the catchment area, where the rainwater is captured. In San Jose Xacxamayo, the rainwater is captured from the households’ roofs, with the majority having concrete roofs. The roof runoff volume of a system can be calculated with the following equation:

Equation (1) Roof runoff volume
1
where:
• Qt = Quantity of monthly rainwater harvested (m3),

• RC = Runoff coefficient,

• R = Total monthly rainfall, and

• A = Roof area or catchment area (m2).

The inflow, Qt, is calculated using the net monthly rainfall, storage catchment size, and the runoff coefficient for the concrete material used and, as adopted for this study, is 0.80 (Biswas & Mandal 2014). To calculate the feasibility of DRWHS for capturing sufficient water for current and future domestic household demand, the water mass balance model was applied based on an approach developed by Fewkes (1999):

Equation (2) Water Mass Balance Equation
2
where:
• Vt= Volume (m3) of rainwater,

• = Volume (m3) of rainwater in storage system at end of monthly interval, t,

• Qt = Inflow (m3) during monthly time interval, t,

• Dt = Demand (m3) during monthly time interval, t, and

• S = Maximum storage capacity.

Time is taken for the monthly storage capacity using the demand of all surveyed users. Each user's current roof size was considered in the calculations. The water balance model was calculated to understand the tanks’ monthly inflow, outflow, and total volume. The performance of DRWHS (reliability) was adapted from Karim et al. (2013) and calculated by taking the total number of users that met their domestic water demand in the monthly time period as described in the following equation:

Equation (3) Reliability of RWHS
3
where:
• Re = Probability of the tank supplying the monthly demand of the users (%),

• = Number of households that did not meet their water demand for that month, and

• N = Total number of users.

## RESULTS AND DISCUSSION

This study provided one of the first integrated assessments of the sustainability of DRWHS, considering social, environmental, technological, and climate factors that affect DRWHS. The results reflect social, water quality, and technical feasibility of DRWHS based on 50 surveys conducted, 17 water analyses, and current and future potential tank storage capacity under a 10%, 15%, and 20% decrease in precipitation for 49 users. Overall, we found that roof sizes were the determined limiting factor of DRWHS to meet year-round water demands for domestic use in a rural, semi-arid area of Mexico. In addition, results of the study suggested that social acceptance, an important factor in DRWHS sustainability, can be achieved with local acceptance of the RWH technology and by acquiring the local skills needed to construct and maintain the systems. In addition, the DRWHS tanks were shown to be economically feasible for many community members to self-finance their own system where existing knowledge and skills are available.

Our study included 100% participation and 100% response to the requests for the 50 surveys. However, some participants did not answer all questions within the survey. Table 1 reports the key socio-economic characteristics of surveyed users. Females made up the majority (96%) of the survey respondents; this was a result of the fact that men were largely unavailable during the time when surveys were carried out. Out of the respondents, 96% identified as one of the heads of household. The average age of the participants was 39 years old and the mean household family size was 4.3 people.

Table 1

Selected socio-economic characteristics of San Jose Xacxamayo

CharacteristicsDescriptionN%
Gender Male
Female 49 96
Age group 18–64 47 42
>64
Household size 1–4 people 31 38
>5 19 66
Occupation Construction 33 66
Agricultural 13 26
Other
Income <$1.25/day 24 48 >$1.25/day 26 52
Potable water No 50 100
Bathroom type Improved 12
Unimproved 42 88
CharacteristicsDescriptionN%
Gender Male
Female 49 96
Age group 18–64 47 42
>64
Household size 1–4 people 31 38
>5 19 66
Occupation Construction 33 66
Agricultural 13 26
Other
Income <$1.25/day 24 48 >$1.25/day 26 52
Potable water No 50 100
Bathroom type Improved 12
Unimproved 42 88

The main economic activity in the majority of the households was construction (66%). Agriculture was second, accounting for 26%, and 8% responded that other sources of income were dominant in their households, such as selling embroidered napkins, weaving palm boxes, selling wood or running a local store. Thus, over half the populations surveyed had at least one household member with knowledge and skills in construction, which are useful to construct and maintain the DRWHS. Monthly income reported (in US dollars) included any government assistance; 64% of participants did receive such aid. Just over half (52%) of the monthly household income of respondents was over $1.25/day/capita. The other 48% reported income levels below the international poverty line of$1.25/day/capita. All respondents (100%) stated they had no access to running potable water in their homes. As related to type of bathroom, 12% of households reported improved sanitation facilities (flush toilets) and the other 88% reported unimproved sanitation facilities (open air and unventilated latrines).

Our results indicated that there were five water sources from which the community could obtain water: wells, a chorro (excess clean water that spills from a tank that transports clean water to a neighboring community), a community cistern of piped water from a nearby source (called the headwater), purchased trucked water, jaguey (community mountain rainwater collection system of 700,000 L capacity implemented in 2005 by the local government), and household RWH collection systems. All 50 surveyed households reported that, for 3 months out of the year, their household does not collect water by foot. The weekly travel frequency to obtain water was answered by 49 out of the 50 respondents: days traveled per week to collect water varied from only 1 day to all 7 days. Most frequently, users travelled once or twice a week to obtain water from local sources. Respondents also noted that most of the weekly travel was done in the dry season (November–April). This result implied that, in the rainy season, most of the surveyed users were able to capture enough rain for household needs and reduce work time required to obtain water from local sources.

All participants had a DRWHS, and only one user did not collect rainwater because of their roof type (woven palm thatch). Participants ranked dependency of their available water sources (jaguey, wells, purchased water, chorro, headwater, and rainwater) in order of reliance and usage for domestic purposes as follows: 26 (52%) respondents ranked rainwater as number 1 (the most used), 18 (36%) ranked it second, 5 (10%) ranked it third, and 1 respondent said they did not use it. Out of all users, 34 respondents reported that they used rainwater for drinking purposes and the other 12 responded that they did not. In terms of the perception of rainwater, the majority of respondents stated that the smell, appearance, and taste were average or above average. One of the users had a bad perception of the rainwater but that user declared that they did utilize rainwater for drinking.

Table 2 shows select characteristics from the existing systems that participants had (age and cost of tank, user contribution to tank construction, and required maintenance). The two types of cisterns that were utilized in the community are concrete and ferrocement. There were more concrete tanks (69) compared to ferrocement tanks (7). Construction of over half of the tanks (55%) was supported by external national and international funds, although a considerable number of participants were financially responsible for their own tank (37%). Construction of all of the ferrocement tanks was supported by external national funds. These types of tanks required more frequent number of repairs at a younger age in comparison to the concrete cistern. Our survey results also suggested that the community did not prefer ferrocement cisterns and that the organizations that promoted their construction did not adequately involve the community members in the planning and implementation process. The lack of social acceptance and limited input of the community in project development of ferrocement tanks could affect the longevity and sustainability of those DRWHS systems. Most households had to contribute with time and/or money regardless of the financial funding source of the tank. The average reported cost contribution by the user per tank was ∼$221, excluding labor. In most cases, the household participated in the construction process by digging the hole and constructing the tank, or by helping with construction. The average number of days for digging the hole needed to construct a concrete cistern (12.6) was higher than that of a ferrocement (2.33), since ferrocement are aboveground structures and only require a minimum of 0.5 m to be dug for the foundation as opposed to 1.9 m for concrete cisterns. The construction process for both averaged about 13 days for each system. Our survey results suggested that length of time to implement a RWH tank was not a determining factor in the social adoptability of these systems. The tank age ranged from <1 year to 18 years, with 38% of tanks having been in existence for over 6 years (all concrete tanks). The number of concrete tanks that required maintenance to fix leaks over their lifespan was low (9%). The average tank age that requires maintenance was 9 years and average total spent on repairs was$50. The ferrocement tank had a higher maintenance requirement (42%) with an average tank age of 1.6 years and an average total spent on repairs of $12.40. The higher repair of the ferrocement tank implied that there could have been inadequate training to the community members in the construction of these types of systems. Out of the 46 who responded to maintenance capability, all reported that their family was capable of maintaining their DRWHS if there were to be a malfunction (i.e. leak). Table 2 DRWHS characteristics CategoryDescriptionConcreteFerrocementTotal Number of cisterns DRWHS 69 76 Government aid No aid 28 28 (37%) International aid 6 (8%) Federal aid 35 42 (55%) Economic costs for construction <$100 24 24
$100–500 14 17$501–1000
>$1000 Household labor contribution (*avg. no. days) Digging 12.6 2.3 Construction 12.3 15 Tank age (years) <1 1–5 35 41 6–10 19 19 11–18 10 10 Maintenance No. of tanks repaired 6 (9%) 3 (42%) 10 (13%) Avg. repaired tank age 1.6 5.3 Avg. total repairs 1.8 1.5 1.6 Avg. repair costs$50 $12$31
CategoryDescriptionConcreteFerrocementTotal
Number of cisterns DRWHS 69 76
Government aid No aid 28 28 (37%)
International aid 6 (8%)
Federal aid 35 42 (55%)
Economic costs for construction <$100 24 24$100–500 14 17
$501–1000 >$1000
Household labor contribution (*avg. no. days) Digging 12.6 2.3
Construction 12.3 15
Tank age (years) <1
1–5 35 41
6–10 19  19
11–18 10  10
Maintenance No. of tanks repaired 6 (9%) 3 (42%) 10 (13%)
Avg. repaired tank age 1.6 5.3
Avg. total repairs 1.8 1.5 1.6
Avg. repair costs $50$12 \$31

### Rainwater quality analysis

Water quality was examined through analysis of physicochemical and bacteriological quality (Table 3). The average water temperature was 19.81°C. The pH ranged from 7.70 to 10.42, with an average pH recorded value of 8.74. In all DRWHS, the pH tended to be more alkaline, likely because of the concrete material on the roofs and tanks. The WHO recommends a pH of 6.5–8.5 for drinking water, however, low alkalinity is not considered a human health hazard (Gould & Nissen-Petersen 1999; WHO 2004; Domènech et al. 2012). The TDS of samples ranged from 23 to 123 ppm with an average of 68 ppm. EC ranged from 33 to 176 μS/cm with a mean of 97 μS/cm. There is no WHO standard for TDS or EC, however, it is generally recommended for TDS to be between 100 and 1,200 ppm for drinking water, of which two of the samples exceeded 100 ppm. Ammonia ranged from less than 0.01 mg/L to 0.245 mg/L with a mean of 0.069 mg/L. Ammonia levels greater than 0.2 mg/L can interfere with the effectiveness of chlorine, and two participants’ tanks reported water with ammonia greater than 0.2 mg/L. Additionally, two household tanks’ results were undetectable, signifying levels lower than 0.1 mg/L. The phosphate ranged from 0.30 mg/L to 0.90 mg/L with an average of 0.55 mg/L. Of the participants, five users’ tanks had undetectable phosphate values, implying levels lower than 0.5 mg/L. The HPC count was high amongst all users’ tanks. Only one user's tank had below 100 CFU/mL at a value of 70 CFU/mL. There are no WHO recommended HPC levels, however, seasonal fluctuations in HPC can indicate maintenance frequency of the user. Four out of the 17 users had coliform bacteria present in their drinking water. Of those with coliform bacteria present, three did not use treatment in the storage tank; one treated with chlorine.

Table 3

Water quality parameters

ParameterRangeMeanWHO guidelines
Temp (°C) 17.1–22.2 19.8 –
pH 7.70–10.42 8.74 6.5–8.5
Conductivity (μS/cm) 33–176 97 –
TDS (ppm) 23–123 68 100–1200
Ammonia (mg/l) <0.01–0.245 0.069 <0.2a
Phosphate (mg/l) 0.3–0.90 0.55 –
Total coliform (CFU/100 ml) 0–500 82
HPC (CFU/ml) 57–10780 1209 –
ParameterRangeMeanWHO guidelines
Temp (°C) 17.1–22.2 19.8 –
pH 7.70–10.42 8.74 6.5–8.5
Conductivity (μS/cm) 33–176 97 –
TDS (ppm) 23–123 68 100–1200
Ammonia (mg/l) <0.01–0.245 0.069 <0.2a
Phosphate (mg/l) 0.3–0.90 0.55 –
Total coliform (CFU/100 ml) 0–500 82
HPC (CFU/ml) 57–10780 1209 –

aRecommended level with chlorine present.

### Technical feasibility

Based on results of the survey, the average domestic weekly water demand of all households surveyed was 1693.17 L, which equates to 56.78 L/c/d (21.4 m3/capita/year). This was below the WHO recommended minimum of 100 L/c/d and just above the WHO poverty threshold 50 L/c/d. Washing clothes and bathing combined made up over half (60%) of household water use.

Figure 1 shows the reliability of the DRWHS systems for both current and future use of the 49 users. A 10%, 15%, and 20% decrease in the current monthly rainfall was taken to project future reliability of the systems under climate change uncertainty. Current and future reliability projections were seasonally variable, with the highest reliability falling in the months of June to September. The variability between different precipitation percentage decrease was not substantial. Overall, in the peak rainy season, a little less than half of the DRWHS users were able to meet their households’ monthly current water demand.
Figure 1

Current and future reliability of DRWHS at San Jose Xacxamayo (considering 1951–2010 monthly rainfall data from the nearest meteorological station, Balcon Diablo A. Texaluca, approximately 9 km distance to San Jose Xacxamayo).

Figure 1

Current and future reliability of DRWHS at San Jose Xacxamayo (considering 1951–2010 monthly rainfall data from the nearest meteorological station, Balcon Diablo A. Texaluca, approximately 9 km distance to San Jose Xacxamayo).

## CONCLUSION

The success and sustainability of DRWHS as a water supply source is highly dependent on factors related to social acceptance, local technical capacity, water quality, and technical feasibility of the DRWHS. Overall, we found that DRWHS were socially acceptable in San Jose Xacxamayo, Mexico, and that water collected was generally fit for human consumption. Our results also suggested that community involvement and contribution in the diagnostics, design, and implementation of water provisions was crucial in order to meet the social acceptance, minimize health risks, and reduce possibilities of the failure of projects that promote DRWHS to increase access to water supply for domestic needs. Our water quality analyses indicated that rainwater could provide quality drinking water as long as additional treatment was performed prior to consumption. There exists a technical feasibility barrier to provide sufficient water supply under current and future precipitation patterns, primarily due to the small average roof size of the participants (70.4 m2), which was not sufficient to capture and store an adequate year round water supply for the average annual demand (87.9 m3). It is recommended to have a roof size of at least 150 m2 for rural, semi-arid communities that aim to rely on DRWHS to supply sufficient quantity of water for annual household domestic demands.

Future research should examine possible heavy metal contamination of DRWHS, to further expand the knowledge and understanding of the role of concrete roofs and storage tanks on rainwater quality. In addition, we recommend future sampling across the year to examine potential seasonal variability. Additionally, more substantial evidence on the economic and social benefits that DRWHS could be garnered through studies that examine conditions before and after their implementation. Future projects would benefit from having regional scale models for small communities to predict future precipitation under different climate change scenarios, as this information could provide for more accurate technical feasibility assessments.

Finally, much of the DRWHS project's success in San Jose Xacxamayo was attributed to the developed trust between the Peace Corps Volunteer and the community, something that is somewhat difficult to quantify. Nevertheless, organizations that implement similar projects in the future should consider this factor, and particularly the influence on social acceptance and feasibility of DRWHS. In addition, for those who do not benefit from an external funding source, DRWHS projects should aim to provide and empower the community members with all the tools that can enable successful implementation.

## ACKNOWLEDGEMENTS

The authors thank the US Peace Corps, the community members of San Jose Xacxamayo, and Puebla's Ministry of the Environment, in particular Ismael Couto, for their participation and contributions towards the rainwater harvesting projects and fieldwork activities. We are also grateful to Dr Ernesto Mangas from BUAP for his help and support with the water analyses. Our gratitude also extends to Marcela León for her assistance in facilitating information exchanges regarding DRWHS through various workshops in San Jose Xacxamayo. The project was funded by USAID, Water Charity, and through private donations.

## REFERENCES

REFERENCES
Adler
I.
Hudson-Edwards
K. A.
Campos
L. C.
2013
.
Journal of Water Supply: Research and Technology – AQUA
62
(
8
),
545
551
.
Alcamo
J.
Henrichs
T.
Rosch
T.
2000
World Water in 2025 – Global modeling and scenario analysis for the World Commission on Water for the 21st Century. Report A0002
.
Center for Environmental Systems Research, University of Kassel
,
Kassel, Germany
.
Arnell
N. W.
2004
.
Global Environmental Change
14
(
1
),
31
52
.
Baguma
D.
Loiskandl
W.
Darnhofer
I.
Jung
H.
Hauser
M.
2010
.
Journal of Water and Health
8
(
2
),
334
345
.
Barron
J.
2009
Rainwater harvesting: a lifeline for human well-being. UNEP/Earthprint, Stevenage, UK
.
Biswas
B. K.
Mandal
B. H.
2014
Construction and evaluation of rainwater harvesting system for domestic use in a remote and rural area of Khulna, Bangladesh
.
International Scholarly Research Notices
2014
,
1
6
.
Brown
A.
Matlock
M. D.
2011
A review of water scarcity indices and methodologies. White paper #106. The Sustainability Consortium, University of Arkansas, Fayetteville, AR, USA
.
Chang
M.
McBroom
M. W.
Beasley
R. S.
2004
.
Journal of Environmental Management
73
(
4
),
307
315
.
Choudhury
I.
Vasudevan
L.
2003
Factors of biological contamination of harvested rainwater for residential consumption
. In:
Hawaii International Conference on Social Sciences
.
University of Hawaii
,
Honolulu, HI, USA
. .
CONAGUA
2012
National Meteorological Service. Secretariat of Environment and Natural Resources of Mexico
. .
Domènech
L.
Heijnen
H.
Saurí
D.
2012
.
Water and Environment Journal
26
,
465
472
.
Doria
M. D. F.
2010
.
Water Policy
12
,
1
19
.
Fewkes
A.
1999
.
Urban Water
1
,
323
333
.
Fewkes
A.
Butler
D.
2000
.
Building Services Engineering Research and Technology
21
(
2
),
99
106
.
Fuentes-Galván
M. L.
Delgado-Galván
X.
Charcas-Salazar
H.
Mora-Rodríguez
J.
Flores
J. L. F.
Benavides
A. C.
2015
Rooftop rainwater harvesting acceptance in three localities of Guanajuato, central Mexico
.
Interciencia
40
(
6
),
403
.
Gould
J.
Nissen-Petersen
E.
1999
Rainwater Catchment Systems for Domestic Supply: Design, Construction and Implementation
.
Intermediate Technology Publications
,
London
,
UK
.
Helmreich
B.
Horn
H.
2009
.
Desalination
248
(
1
),
118
124
.
Imteaz
M. A.
Adeboye
O. B.
Rayburg
S.
Shanableh
A.
2012
.
Conservation and Recycling
62
,
51
55
.
INEGI
2010
Censuses of population and housing
. .
Jenkins
D.
Pearson
F.
Moore
E.
Sun
J. K.
Valentine
R.
1978
Feasibility of rainwater collection systems in California
.
Contribution No. 173
,
Californian Water Resources Centre, University of California, Riverside, CA
,
USA
.
Karim
M. R.
Rimi
R. A.
Billah
M. S.
2013
Reliability analysis of household rainwater harvesting tanks in the coastal areas of Bangladesh using daily water balance model
. In:
20th International Congress on Modelling and Simulation
.
Adelaide, Australia
, pp.
2639
2645
.
Liaw
C. H.
Tsai
Y. L.
2004
.
Journal of the American Water Resources Association
40
,
901
912
.
Lizárraga-Mendiola
L.
Vázquez-Rodríguez
G.
Blanco-Piñón
A.
Rangel-Martínez
Y.
González-Sandoval
M.
2015
.
Water
7
(
9
),
4622
4637
.
Meera
V.
Ahammed
M. M.
2006
Water quality of rooftop rainwater harvesting systems: a review
.
Journal of Water Supply: Research and Technology – AQUA
55
,
257
268
.
Mendez
C. B.
Klenzendorf
J. B.
Afshar
B. R.
Simmons
M. T.
Barrett
M. E.
Kinney
K. A.
Kirisits
M. J.
2011
.
Water Research
45
(
5
),
2049
2059
.
Mun
J. S.
Han
M. Y.
2012
.
Journal of Environmental Management
93
(
1
),
147
153
.
Mutekwa
V.
Kusangaya
S.
2007
.
Water SA
32
(
3
),
437
444
.
Neibaur
E.
2015
Sustainability Analysis of Domestic Rainwater Harvesting Systems for Current and Future Water Security in Rural Mexico
.
Master's Thesis
.
Florida International University, Miami, FL, USA
.
Ndiritu
J.
Odiyo
J. O.
Makungo
R.
Ntuli
C.
Mwaka
B.
2011
.
Hydrological Sciences Journal–Journal des Sciences Hydrologiques
56
(
2
),
238
248
.
Organisation for Economic Co-operation and Development (OECD)
2013
Improving Education in Mexico: A State-Level Perspective from Puebla, OECD, Paris, France
.
Palla
A.
Gnecco
I.
Lanza
L. G.
2011
.
Journal of Hydrology
40
,
65
76
.
Pushpangadan
K.
Sivanandan
P. K.
2001
Technology, quality and cost of water from DRWH – a case study in Kerala
,
India
.
Proceedings of the Rainwater Harvesting Conference, Delhi, India, E 3.1–E 3.9
.
Schets
F.
Italiaander
R.
Van Den Berg
H. H. J. L.
de Roda Husman
A.
2010
.
Journal of Water and Health
8
(
2
),
224
235
.
United Nations (UN)
2007
Coping with water scarcity: challenge of the twenty-first century. 2007 World Water Day. UN-Water. FAO, Rome, Italy
.
United Nations (UN)
2012
The Millennium Development Goals Report. New York, NY, USA
.
Wallace
C. D.
Bailey
R. T.
2015
.
JAWRA Journal of the American Water Resources Association
51
(
1
),
185
199
.
Ward
S.
Memon
F.
Butler
D.
2010
.
Water Science and Technology
61
(
7
),
1707
1714
.
WHO
2004
Guidelines for Drinking-water Quality
.
World Health Organization
,
Geneva, Switzerland
.
WHO
2008
Guidelines for Drinking-water Quality-Incorporating First Addendum. Recommendations
, 3rd edn.
Vol. 1
.
World Health Organization
,
Geneva
,
Switzerland
.
Youn
S. G.
Chung
E. S.
Kang
W. G.
Sung
J. H.
2012
.
Resources, Conservation and Recycling
65
,
136
144
.