1. Introduction

The Millennium Development Goals (MDGs), developed in 1990, established a number of targets to be reached by the year 2015. Target 7c was to halve the number of people without reliable access to safe drinking water. Since the 1990s, much progress on this target has been made, particularly in urban areas (Mukungu 2000; United Nations 2014). From 1990 to 2010, access to “improved” water in rural sub-Saharan Africa increased from 36% to 49%, which means falling short of Target 10 by 30% in these areas (Bradley and Bartram 2013; WHO/UNICEF 2012).

The MDG target highlights two health-related dimensions of drinking water: quality and quantity. For MDG monitoring, an improved water source is defined as one “that, by the nature of its construction and when properly used, adequately protects the source from outside contamination, particularly faecal matter” (http://www.wssinfo.org/definitions-methods/watsan-categories/). However, there is a push to move beyond this paradigm to monitor access in more clear categories, by measuring (i) those using drinking water directly collected from surface water (e.g., ponds)—unimproved water sources; (ii) those using improved sources other than piped household connections; and (iii) those with household connections in a dwelling, plot, or yard (WHO/UNICEF 2012). These categories have implications for potential contamination but also for the quantity of water. Specifically, surface water sources are highly prone to fecal and other contamination (Howard et al. 1999, 2003; Pedley and Howard 1997) and evaporation (Reimann et al. 2003), whereas piped connections are typically drawn from groundwater sources, which tend to be more reliable (do not evaporate) and less contaminated (Pearson et al. 2008). Piped sources in dwellings or within plots also mean that households can more easily fetch water. Studies indicate that water-related diseases dramatically decrease when piped sources are located in a dwelling or on a plot (Bartram and Cairncross 2010).

The Joint Monitoring Programme (JMP), conducted through WHO and UNICEF, monitors and reports improved access to adequate water to assess progress in meeting the MDGs and future targets. While these monitoring efforts have been critical to understanding changes in access over time, we still lack a realistic understanding of day-to-day access to adequate water for most of the world. What is clear is that an astounding 748 million people still lack access to improved water (WHO/UNICEF 2014). Even with such a large swathe of society living with unimproved access to water, very little is known about how households cope with poor quality and/or low quantities—or water scarcity. The dimensions of access to adequate water are the focus of the research here and can be usefully understood in relation to the notion of water security, which is a concept put forward as a focus for post-2015 sustainable development goals.

Water security has been defined as “…sustainable access to adequate quantities of acceptable quality water for sustaining livelihoods, human well-being, and socio-economic development, for ensuring protection against water-borne pollution and water-related disasters, and for preserving ecosystems in a climate of peace and political stability” (http://www.unwater.org/other-resources/for-the-media/allpressreleases/water-security/en/; United Nations 2012). This definition includes aspects of both quality and quantity and underscores several factors that influence water security, including the physical environment and political systems. The current WHO/UNICEF monitoring program does not account for or measure dimensions of water security including seasonal fluctuations in availability, maintenance and governance issues that affect the usability of or access to water sources, or the dynamic nature of changing levels of reliability in water source access.

One of the major global factors in changing water security is climate change. Climate change estimates for sub-Saharan Africa vary by region, with differing estimates of freshwater availability into the next decades. A recent review of climate change science states that in addition to the clear impacts of water for household hygiene and drinking, changes in water availability will likely impact crop yields and disease vectors, both of which will affect health in the region (Field et al. 2014). Still, both the uncertainty in climate change models and the inherent microclimatic conditions make it difficult to predict changes in precipitation and land temperature (which affect evaporation rates for surface water sources) (Boko et al. 2007). Even slight changes in precipitation in the arid and semiarid regions can lead to dramatic changes in runoff and resulting surface water availability (Fekete et al. 2004). Attributing water scarcity to climate change alone is myopic. In fact, a suite of issues act upon water availability including sedimentation and land-use practices, contamination of sources, governance, and population pressures. Even in the absence of climate change, current trends in population and water use indicate that before 2025 more African countries will exceed the limits of their land-based water resources (Ashton 2002). Climate change is expected to exacerbate scarcity, with estimates as high as 75–250 million people at risk by the 2020s (Bates et al. 2008).

The research here draws from the concept of water security to examine one dimension—water adequacy—and related coping strategies in the face of scarcity in rural communities in the semiarid savannah, which have undergone significant livelihood and water access changes as a result of national policy changes. Specifically, the aims of this research were to 1) understand aspects of water scarcity experienced in three rural villages in southwestern Uganda, 2) examine differences by demographic groups and type of source, 3) assess relationships between different factors related to water access, and 4) explore the coping strategies used during times of scarcity by those of varying ethnicity or livelihood. In the subsequent discussion, health implications are then highlighted, and relevant lessons for the future between climate change and estimated exacerbation of water insecurity are discussed.

2. Methods

3. Results

The survey was conducted in 92 households out of the original 100 enrolled in the three communities. Eight households were lost to follow-up due to relocation outside of the three study communities; there were no refusals. Most mothers and fathers in households did not attend any secondary school (Table 1). About half of the respondents reported that their primary livelihood was derived from agricultural activities (52%), followed by cattle, fishing, and other activities. The average age of the acting head of household was 45 years and most households self-reported their level of wealth as moderate or average (54%) followed by low to very low wealth (26%). Overall, 21% of households reported experiencing inaccessibility of their primary household water source in the last two weeks (during the midpoint of the dry season), and 61% report that this primary source is temporary or seasonal.

Table 1.

Characteristics of survey respondents by seasonality of primary household water source.

View Table

In comparing the household demographics and water source characteristics between those relying on a primary water source that is reported to be permanent versus temporary (Table 1), most characteristics were similar between the two groups. However, households accessing temporary sources tended to have higher levels of at least primary education for the father (24% vs 10% for those accessing permanent sources) and mother (33% vs 6%), to be of Bahima ethnicity, to use cattle or agriculture as the primary livelihood, and rely on rainwater (52%) followed by farm pond (45%) as the type of water source. Surprisingly, those accessing permanent sources reported water inaccessibility within the last two weeks (29% vs 16%). As expected, those accessing permanent sources tended to use a borehole or the lake.

Assessment of the relationships between different factors related to water access, using Spearman’s rank correlations (Table 2), indicated that self-reported level of better access to water was significantly correlated with fewer minutes spent walking to the water source (Spearman’s rho = 0.585; 95% CI 0.432–0.705). The self-reported better access was weakly correlated with access to a more public or shared source, although at a marginal significance level (Spearman’s rho = 0.203; p = 0.0519). Surprisingly, more public ownership was also weakly correlated with experiencing inaccessibility at least once in the last two weeks of April (Spearman’s rho = 0.209; 95% CI 0.002–0.399). During the preceding two weeks, 40% of those using boreholes experienced difficulty in accessing their primary source, indicating a maintenance issue with one of the two public boreholes and in the study area. In addition, 28% of households using rainwater experienced difficulty in accessing their primary source.

Table 2.

Comparison of measures of water insecurity and access. The items in bold font significant at the p < 0.05 level.

View Table

In the wet season, there was a fairly even split between reliance on boreholes, rainwater, or farm ponds (about 30% each). In the dry season, an increased number of households (50%) rely on boreholes, followed by lakes, farm ponds, and government-owned large ponds. Regardless of season, those accessing a borehole tend to use the source more consistently. In contrast, those accessing smaller surface water sources (farm ponds) in the wet season tend to shift to larger surface water sources in the dry season, such as lakes, rivers, and particularly government-owned large ponds. Those who use rainwater in the wet season rely predominantly on boreholes followed by lakes in the dry season.

To cope with water scarcity, Bahima, the relatively wealthy pastoralists, reported primarily relying on either legal migration (38%) or sharing (33%) (Table 3). This is because they typically have kinship and social networks with other pastoralists in other locations. Bairu, the agriculturalists, tend to rely on sharing (37%) and purchasing water (28%). Some Bairu also report illegal migration (7%). Although households of other ethnic groups are few in this study (n = 14 are nonnative to this region), these households reported reliance on legal migration (36%) and purchasing water (21%). This group also reported the highest percentage of no coping strategies used (29%). These groups may be less able to share with neighbors because of the lack of social connections.

Table 3.

Water access characteristics by primary coping strategy used during times of scarcity.

View Table

Households reporting high wealth status tend to share water (n = 18) as do the poorest (n = 24), indicating that the wealthy may be sharing with the poorest households. This compares to slightly lower levels among the middle income group (30%). Households reporting low income reported the highest level of illegal migration (8%). The next highest was reported by wealthy households (6%), which may indicate that some poor households are more limited to tenuous coping strategies and that any fines or bribes are not a deterrent for the wealthy group. The middle income group had the highest reliance on purchasing water during times of scarcity (30%), followed by the poorest households (25%). It appears that migration, whether legal or illegal, remains a well-used coping strategy in these communities, despite land privatization and the creation of Lake Mburo National Park.

Households with very high self-reported access to water relied on illegal migration as a coping strategy (n = 11). Those in the moderate access group utilize every coping strategy category, with a large proportion sharing (33%) followed by purchasing water (31%). Households reporting low (n = 2) and very low access (n = 1) were few. These households either report no coping strategies or sole reliance on purchasing water. The burden of paying for water may be the reason for reporting very low access to water for this household.

Households using public water sources (n = 57) tended to share (35%) or purchase water (25%). This group was also the only group to report illegal migration (7%). Those using a shared (n = 25) or public (n = 9) source tended to rely on legal migration during scarcity. Those using a personal source reported the highest percentage of no coping strategies (22%), followed by migration (34%) and sharing (33%).

4. Discussion

In this study, almost half of the participating households located in southwestern Uganda access a household water source that becomes unusable in the dry season because of evaporation. Of those accessing “permanent” sources, about 30% experienced inaccessibility of their primary household water source within the last two weeks of April (during the midpoint of the dry season). This inaccessibility of a permanent water source was because of a malfunctioning borehole in one community, as observed during the process of data collection. Access to other sources considered permanent, such as the lake, were more reliable over the preceding two weeks; however, surface water sources have previously been shown to be highly contaminated in this area (Pearson et al. 2008). These findings suggest separate long- and short-term causes of insecurity. A short-term cause of insecurity is the lack of maintenance or proper management of water sources by the water management committee. Two long-term sources of insecurity that are evident include the lack of government provision of sustainable water supplies (e.g., boreholes) and the inherent annual water scarcity related to living in the semiarid savannah.

Table 4 illustrates the seasonal diversity of water sources utilized by households. The most likely sources of relatively safe water are boreholes, and it is interesting that, in this environment, during the dry season borehole water is used by 50% of households (46 of 92), whereas borehole use falls to 32% in the wet season. Meta-analyses of coliform counts on diverse water sources (Bain et al. 2014) suggest that harvested rainwater is often contaminated, and our data show that farm ponds in the area have marked coliform contamination (Pearson et al. 2008), so that, contrary to what might have been expected, use of polluted sources appears to rise in the rainy season. The data on seasonal use of sources point to a need for education on choosing safe sources in the wet season and the need for more studies of harvested rainwater quality in the study area.

Table 4.

Numbers of households using various water source types in wet vs dry season. Boldface italic numbers indicate a change in source type across seasons. Regular-font numbers on the diagonal indicate the same source type across seasons. Here, GB pond indicates government big pond.

View Table

For households, self-reported better access to water was associated with proximity to the source and, to a lesser degree, with using more public or shared sources. Those without access to public sources tended to utilize migration as the primary coping strategy. There was also a subset of wealthy Bairu with access to public sources. When those sources became inaccessible, the wealthier resorted to illegal migration to cope. This study also identified a high level of water sharing and reciprocity (although at much lower levels for outside ethnic groups) between wealthy and poor alike. This most likely indicates a strong reliance on social capital bridging between the wealthy and poor for coping (Bisung and Elliott 2014) as wealthy households tended to have larger water sources (because of having larger plots of land) that persisted longer into the dry season. Because outside ethnic groups appear partially excluded from this coping mechanism, with higher reliance on migration and purchasing water, this suggests that there may be aspects of strong social capital, which are beneficial for Bahima and Bairu but may negatively impact outsiders. For example, social norms defined by these ethnic groups may make it difficult for other ethnic groups to assimilate, to participate, and to form social bonds that are used to cope in times of water scarcity, as suggested elsewhere (Bisung and Elliott 2014). The observed social aspects of water access and coping strategies are entrenched in a much longer history of resource access and ethnic privilege, which existed prior to colonialism but was exacerbated by a series of geopolitical processes over the past 150 years [for a historical account, see Pearson and Muchunguzi (2011)]. Whereas historical power struggles existed between the Bairu and Bahima, these may have diminished in comparison with the exclusion of those from outside ethnic groups.

Other studies evaluating coping strategies and adaptation in arid lands tend to focus on livelihood adaptation of pastoralists (Huho et al. 2011) or agricultural and food strategies (Rufino et al. 2013) rather than explicitly focus on strategies used to access household water. However, an exception is a study in Nigeria that identified that households tended to “close” sources to outsiders during times of scarcity and, echoing our study’s findings, there was thus a reliance on purchasing water (Adeniji-Oloukoi et al. 2013). Notable differences between the present study and the Nigerian study are that the sample in Nigeria primarily relied on wells and boreholes that are less prone to evaporation and contamination. Migration was not an identified coping strategy (presumably because of the lack of pastoral participants), and a series of behavioral coping strategies used to conserve water were highlighted. These differences highlight important considerations for future water insecurity research and potential policy implications.

First, the lack of behavioral water conservation strategies in this study indicates that households tend to minimize water usage and inhabit a water insecure environment. This is indicated by the reliance on surface water sources and a high level of recent experienced insecurity. This means that households must find alternative sources rather than alter water usage activities to cope. Locating alternative sources involves sharing, migration (both legal and illegal), and/or purchasing water. Because of the historical reliance on migration in this region, this remains a primary coping strategy for many households, despite government efforts to promote sedentarism (Cook 1994). However, it is unknown whether the driving force behind migration is the needs of livestock or humans but it is likely to be both.

Second, the continued reliance on migration (including illicit migration) suggests the need for local or national government consideration of water needs for both household and livestock consumption. Allocation of large surface water sources on public land may be a reasonable action to reduce illicit migration (particularly for those with herds of cattle and if grazing is permitted on such land). District-level government could also prioritize allocation of geologic and hydrologic surveying, with the subsequent provision of groundwater drinking water sources to reduce the reliance on surface water sources for household consumption. Actions such as these may reduce migration and also potentially high levels of contamination associated with surface water sources in the area (Pearson et al. 2008). Policies that integrate the importance of adequate water for households and for livestock may be more effective and sustainable.

This study has limitations: First, water used to support livelihoods (e.g., livestock or crops) was not assessed. The focus of this research was household water. However, without access to water for food or marketable goods, households still experience water scarcity (and food insecurity), with severe health implications. Often, water sources are shared between humans and livestock (Pearson et al. 2008) and thus water availability in one domain affects the other. Any policy or intervention to address household water must also consider the water needed for livelihoods in order to sustainably address water availability. Second, the self-reported data inherently involve potential bias in recall and reporting bias. This is particularly salient in terms of reporting illegal migration. Participants may have been concerned about the consequences of reporting this coping strategy and thus there is likely an underreporting here. Third, households may rely on more than one source type in each season. This study reported only the primary source type for wet and dry seasons. Last, the cross-sectional nature of this study means that comparison of recent scarcity in the wet and dry seasons was not possible. Future studies could benefit from collecting objective data on water-use quantities, water access for livelihoods, and longitudinal data on experiences of scarcity or inaccessibility of water.

5. Conclusions

This study underscores how access to water resources, particularly in times of scarcity, involves social systems and governance. The absence of diversity in coping strategies (or the uniformity of coping strategies) by the poorest households and reliance on purchasing water among the middle income group have the potential to exacerbate socioeconomic health inequalities. The lack of adequate water for drinking and hygiene has been associated with a suite of infectious diseases (Pruss et al. 2002). With changing water availability expected across the planet, water scarcity will be a new phenomenon in some locations and exacerbated in others. The identified long- and short-term causes of scarcity in this study could partially be mediated by proper maintenance and governance of water provision. Any policies or actions to diminish water insecurity must foster such trust and reciprocity upon which many households already rely. Specific interventions or policies must also be considered for migrants and the poorest households (both economically poor and poor in social capital).