A growing concern: investigating sustainable water quality and urban food production in Freetown, Sierra Leone Open Access

EPA

Environmental Protection Agency

FAO

Food and Agriculture Organization of the United Nations

GV

guideline value

HDI

Human Development Index

IAEA

International Atomic Energy Agency

IPE

International Plant Analytical Exchange

OECD

Organization for Economic Co-operation and Development

TDS

total dissolved solid

UPA

urban and peri-urban agriculture

UNDP

United Nations Development Programme

WHO

World Health Organization

XRF

X-ray fluorescence

Cities, and their rapidly expanding urban ecological footprints, are growing faster in sub-Saharan Africa than in any other region in the world. Estimates suggest that Africa's population will double between now and 2050, and two-thirds of this population increase will be absorbed by urban areas (OECD 2020). As urban resources are stretched like never before, there is an urgent need for innovative, inclusive, interdisciplinary research that seeks to understand the challenges faced by urban citizens in Africa, as well as potential solutions that will pave the way for healthier and more sustainable urban living. One spontaneous, community-driven response to the interlocking problems that poor city-dwellers face has been the proliferation of urban and peri-urban agriculture (UPA), often taking place at sites highly contaminated by industrial wastewater and sewage. Recent research carried out across Africa suggests that UPA not only provides the urban poor with important employment and food provisioning opportunities, but it can strengthen the social fabric of communities, provide significant benefits for women, and, if supported, provide important ecological functions to address the growing environmental footprints of cities (Eigenbrod & Gruda 2015).

Sierra Leone is currently ranked as one of the poorest countries in the world (HDI rank 182 out of 189), with a range of very weak economic and social development indicators (UNDP 2022). The rapid expansion of the country's capital, Freetown, partly caused by an influx of internally displaced migrants during the country's decade-long civil war (1991–2001), has left a considerable proportion of the urban population without reliable income or infrastructure for livelihood support. As the urban population has continued to grow rapidly, an accompanying decline in rural food production has meant that Sierra Leone is now only able to produce 20% of the annual domestic requirements of rice. This has left Freetown's population particularly vulnerable to food insecurity (Lynch et al. 2013), and the situation has been further exacerbated by world events causing dramatic spikes in global food prices. To further compound these challenges, urban pollution and malnutrition in Freetown have also increased dramatically in recent years.

This study investigates the relationship between soil and water quality and corresponding food security and public health in Freetown and, ultimately, seeks to contribute an interdisciplinary perspective by focusing on an under-researched area: the intertwined health and environmental impacts of UPA on local communities. Addressing this knowledge gap seems imperative, as most major UPA sites in Freetown are located within wetlands, on riverbanks and/or along small streams; crop cultivation often takes place in dried-out riverbeds during the dry season. The water from streams and shallow groundwater wells used for irrigation at UPA sites is frequently contaminated by domestic and industrial wastewater, and these may be the only reliable water sources available. Furthermore, it is currently unknown what levels of pollution are being transported through the food chain of Freetown's UPA sites, placing already precious food and water supplies even more at risk (Conteh et al. 2017). Against this backdrop, our study aimed to (i) monitor and characterise the level of water, sediment and crop contamination at three distinct UPA sites in Freetown; and (ii) identify UPA community members' concerns regarding environmental, livelihood, health and wellbeing issues. Results provide the much-needed data on (i) the connected health, socio-economic and environmental effects of urban food production on local communities in low- and middle-income regions and (ii) the potential for nature-based strategies such as phytoremediation to address these critical issues.

Drawing upon a mixed-method, interdisciplinary approach, the fieldwork which informs the paper adopted methodological tools from the social sciences, engineering, and health sciences. Initially, Participatory Research Appraisal was undertaken with the assistance of Freetown urban farmers themselves, to map the diverse characteristics and challenges at the UPA sites and ensure that co-production of local knowledge featured prominently in our research. This was complemented by stakeholder interviews, environmental monitoring and sampling, laboratory analysis of water, sediment and crop samples, and socio-economic surveys at the study sites.

River water was monitored at the three UPA sites once per month between November 2020 and January 2022. Conductivity, pH, total dissolved solids (TDSs), turbidity and water temperature were measured in-situ using calibrated portable multi-meters (Thermo Scientific™). Chemical metal and nutrient concentrations were measured on site using a portable heavy metals analysis system and digital multiparameter photometer (Wagtech™). Additionally, raw water samples were obtained from each site and tested for presumptive thermotolerant faecal coliforms (including Escherichia coli) at the Sierra Leone National Water Quality Laboratory.

Sediment samples (∼50 g), which were taken in coordination with water sampling, were obtained from parts of the riverbed used for farming during the dry season. Following collection, samples were transported to the University of Sierra Leone Fourah Bay College Physics Department laboratory, where they were dried and sub-sampled (20 g) for chemical analysis using a Thermo Scientific™ Niton™ XL3t GOLDD+ X-ray fluorescence (XRF) analyser with International Atomic Energy Agency (IAEA) reference material CH-1 (marine sediment) for calibration.

Crop biomass was analysed using a similar methodology to the sediment sampling. Sweet-potato leaves were investigated for this study, as this crop is common to all three sites. Since the leaves are commonly consumed (and re-harvested repeatedly throughout a single growing season) and sweet potato is known to accumulate more metal pollutants in its leaves than other local crops, contaminants may be more clearly identified in its leaf tissues (Thullah et al. 2019). Between January and June 2021 and from January to February 2022, sweet-potato leaf samples were obtained monthly (in coordination with water and sediment sampling) from the study sites and then transported to Fourah Bay College for metals analysis by XRF, using International Plant Analytical Exchange (IPE) reference material for Potato (mixture)/Solanum tuberosum (WEPAL-IPE-971) for calibration. In addition to quantification of chemical concentrations of the plant biomass, the risk to human health by the intake of heavy-metal-contaminated sweet-potato leaves was characterised by comparing the tolerable daily intake (TDI) of each heavy metal detected against the average human body mass, which has been estimated at 60.7 kg for the African continent (Baars et al. 2001; Walpole et al. 2012).

Supporting socio-economic surveys were carried out in February 2022 with farming communities at the three UPA sites. For these surveys, participants were selected at each UPA site using a ‘snowball’ sampling method and interviewed for roughly an hour each. Snowball or ‘chain referral’ sampling is a non-probability sampling method where research participants recruit other participants as part of the sample. In this study, it was a useful way to conduct research on a population with specific characteristics (knowledge of urban farming in Freetown). In total, there were n = 75 interviewees (25 per site), who answered questions about household demographics, water-collection and sanitation habits, health status and farming methods. An end-of-project workshop, bringing together project researchers, farmers, local government officials and other stakeholders, was held in Freetown in May 2022 to discuss the results and paths going forward.

Water: Survey respondents revealed there is a water shortage in Freetown during the dry season, and the lack of access to safely managed drinking and irrigation water by UPA farmers is a critical and ongoing issue. Over 80% of crop irrigation water came from untreated sources at all three of the UPA study sites, with river water being used for 25%–85% of irrigation. Less than 5% of respondents used river water for drinking. Water quality monitoring shows that the river water used for irrigation at all three UPA sites failed to meet the WHO Guidelines for Drinking Water (2017) for turbidity, Cr, Mn, phosphate, faecal coliforms and E. coli. Relative to guideline values (GVs) for irrigation water (Table 1), water samples exceeded GVs for As, Mn, Cr, E. coli, turbidity, and pH. The toxic heavy metals Cd, Hg, Ni, and Pb were undetected. With respect to nutrients, peak nitrate and K concentrations were observed during the rainy season while maximum phosphate concentrations occurred during the dry season. Survey results suggest that seasonal variation in the types of fertiliser applied at the UPA sites may be a contributing factor, as well as excess runoff of fertiliser during extensive rain events. Extremely high readings for total faecal coliforms and E. coli were observed in river samples from Kingtom and New England, with measurements above 10⁵ CFU per 100 mL. Regent exhibited relatively less contamination by presumptive faecal coliforms at 10² CFU per 100 mL. Survey results indicate that these extreme levels of faecal pollution may be attributed to the flooding of latrines during the rainy season and the common practice of emptying latrines into the river and/or using the river for open defaecation.

Table 1

View Large

River water was found to be used for irrigation by 20% of the farmers surveyed at Regent, 55% at New England and >85% at Kingtom. Correspondingly, 60%, 92% and 100% of respondents at Kingtom, New England, and Regent, respectively, suffered from skin infections, indicating a possible causal relationship between using untreated water for irrigation and skin infections. Respondents from Regent, who consumed water from many different types of improved and unimproved water sources (with 84% disinfecting their water), reported no incidents of diarrhoea for the previous four weeks before the survey. Conversely, in Kingtom 68% of respondents disinfected their drinking water but 32% still suffered from diarrhoea in the previous four weeks before the survey. Microbial analyses showed much lower counts of presumptive faecal coliforms and E. coli for Regent surface water; however, survey results also show a higher rate of Regent respondents disinfecting or boiling their water before consumption. Hence, respondents consuming potable water may be less likely to disinfect it, based on the assumption it has been properly treated.

Sediment: Of the 30 metals analysed in the sediment samples, those with significant presence included (in descending order): Fe > Al > Mg > Mn > Cu > Si > Zn > Co > Cr > Pb. The other metals analysed (As, Ag, Au, Bi, Cd, Cs, Hg, Mo, Nb, Ni, Ru, S, Sn, Sr, Th, Ti, V, W, Zr) had negligible presence. Concentrations and corresponding GVs are presented in Table 2; in the absence of specific West African regulations, GVs are based on general recommendations for metals in agricultural soil to support crop health.

Table 2

View Large

Table 2 results show that Cu sediment concentrations consistently exceeded recommended or regulatory values (RVs) for agricultural soil. Concentrations of Zn, Co, Cr, Fe, and Mn were elevated as well. Natural deposits of Cr and Co are typical of laterite soil; this may be a cause of their high concentrations in the sediment samples. Freetown sits on a peninsula characterised by laterite topsoil with high Fe, Mn, and clay content (ISDA 2021), which may contribute to the Fe and Mn abundance in the samples. Laterite soil is frequently unsuitable for agriculture due to poor nutrient content, and the high Fe content of West Africa's soil has long been an issue for crop production (Kumar et al. 2022). It is also possible that metals may be accumulating in the UPA soil due to improperly managed runoff from Freetown's substantial industries. Kamara & Thullah (2021) observed similar levels of Cr, Cu, Fe, Pb, and Zn in the soil around the Kingtom and Granville Brooke dumpsites in Freetown, though their measurements for Zn and Pb were considerably higher (1,350 and 434 mg/kg vs 356 and 1 mg/kg in this study, respectively); this may be due to the closer proximity of their study locations to the dumpsites. A comprehensive study by Conteh et al. (2017), which included a greater range of UPA sites over a shorter time-period, found similar but lower soil metal concentrations which remained below FAO/WHO RV.

Crop biomass: Samples of sweet-potato leaf were measured for the following metals, listed in order of abundance: Al > Cu > Fe > Ni > Pb > Cr > As > Cd > Hg, as shown by values in Table 3. Relative to RV for leafy or root food crops (WHO 1996; WHO/FAO 2019), all metal concentrations were found to exceed regulatory limits, except for Fe and Al (not regulated) and Hg (not detected). Many of these regulations are critical as direct, chronic consumption of these metals can be dangerous in elevated quantities. Ni and Cu were found in concentrations several times higher than the RV; As, Cr, Cd, and Pb levels are even more concerning, with concentrations up to >100 times higher than the RV (Table 3). Crop metal concentrations were consistently lower (though still exceeding RV) at Regent; this is pronounced for Cr, where mean values were >100 times above the RV at Kingtom and New England, but only 14 times higher at Regent. These findings indicate that much of the crop metal-contamination may be caused by industrial/dumpsite leachate at the more urban sites (Kingtom and New England).

Table 3

View Large

Social implications: The social survey results provide further perspective, placing the environmental water, sediment and crop results into the context of broader issues and behaviour. Most of the household and farmer survey interviewees were women, with 60% in Kingtom, 88% in New England and 100% in Regent, in line with related studies (Cadzow & Binns 2016). An investigation into the roots of water source contamination requires an examination of the sanitation facilities used. Most of the land used for UPA cultivation in Kingtom and New England was either part of a riverbed, floodplain and/or natural wetland, all of which are affected by flooding during the rainy season, thereby reducing available space for farming even further. Furthermore, in the absence of any alternative measures, urban streams in Freetown are regularly used for liquid and solid-waste disposal, which means that UPA sites are heavily impacted by faecal matter and solid-waste pollution. Only 29% of interviewees reported using improved sanitation services (e.g., ventilated improved pit latrines), considerably less than previous estimates (68% from Tumwebaze et al. 2022) and the WHO/UNICEF assertation that 83% of urban areas in Sierra Leone have adequate sanitation facilities (WHO 2010). Accordingly, there are few safe options for latrine waste disposal, as evidenced by the extremely high levels of faecal coliforms and E. coli measured in river irrigation water (Table 1). Interviewees in Kingtom confirmed that much of their sewage waste ends up being dumped in the river (44%). In New England, 12% said their latrines were emptied by a septic tanker, while 56% stated that the waste was removed ‘manually’, which suggests that at least some of this waste may end up being disposed of in the river. The situation in Regent seems improved, with 40% responding that latrines were emptied by a septic tanker; however, 60% said that once a latrine is full, it is simply closed and a new one is built. In addition to being unsustainable in the long term, this practice poses an inevitable risk of contaminant leaching from old, leaking latrines into local river and groundwater.

Going forward: These baseline data indicate that pollutant concentrations in water, sediment and crops at the Freetown UPA study sites exceed regulatory limits (Tables 1–3), with sweet-potato leaves accumulating heavy metals to concentrations exceeding WHO limits for food (Figure 3). Microbial contamination of irrigation water has also been identified as a significant issue (Table 1).

Protection of the water bodies in Freetown (apart from providing the infrastructure needed for functional liquid and solid-waste disposal services) could be enhanced by enforcement of the existing Sierra Leone Wetland Policy (Government of Sierra Leone 1970) and of the urban wetland protection and construction ban near water bodies, flood plains and urban farming sites. According to the Wetland Policy (Government of Sierra Leone 1970), most of the low-lying wetland areas unsuitable for construction purposes are owned by the local government or the state, in order to allow local people to farm crops (Lynch et al. 2013). Unfortunately, this policy is not well known among UPA farmers and local politicians and is thus not enforced; observations conducted during this research showed a critical decline in available land for farming due to increased construction activities at all three study sites.

Nature-based solutions, such as riparian filter or phytoremediation barriers, may also be a viable option for protecting Freetown's urban water bodies, while also contributing to the enhancement of water quality and food security. While phytoremediation is considered to be one of the most sustainable water treatment options for developing countries, it remains largely untested for use in low-income UPA (Anning et al. 2013). Riparian (vegetated) filter strips are primarily used in agriculture to control non-point-source pollution and capture sediments from storm water runoff; they have been found to be highly effective in reducing the amount of nutrients and pesticides in agricultural runoff (Lovell & Sullivan 2006; Prosser et al. 2020). In addition, vegetated filter strips could potentially prevent people from littering and/or other activities close to water bodies due to limited access.

Constructed wetlands function through the process of phytoremediation. Phytoremediation is defined as a group of plant-based remediation methods to reduce, remove, break down or immobilise environmental contaminants (Purakayastha & Chhonkar 2010). Additionally, as plant roots anchor themselves to soil in support of above-ground biomass, a reinforced soil matrix is produced that is less prone to shear failure, thereby creating the additional benefit of increased soil stability and reduced erosion (Meijer et al. 2018; Krzeminska et al. 2019). Anning et al. (2013) show that Ghanaian wetland plants can sufficiently remove heavy metals from wastewater; several of these plants including Typha latifolia (cattail) and Thalia geniculata (arrowroot) are also indigenous to Sierra Leone riverine systems (FAO 2019). While phytoremediation is considered one of the most sustainable water-treatment options for developing countries, it remains largely untested for use in low-income UPA (Anning et al. 2013). Thus, successful low-cost green infrastructure aimed at reducing the exposure to heavy metals in UPA could potentially have a significant, direct and tangible impact on the health and wellbeing of farmers and end-users of their crops.

Baseline study results presented here highlight the scope of the problem at Freetown UPA sites. As part of an ongoing follow-up study, phytoremediation barriers were planted at the three sites in the spring of 2023, using indigenous hyperaccumulator plants including Leucaena leucophala (mimosa tree), Cana indica (Sierra Leone arrowroot, or ‘flower of the dead’) and local grasses. These barriers were planted at all three study sites along the perimeter of crop beds and nearby river banks. Monitoring is currently underway to evaluate the influence of these barriers on mitigating food crop contamination. Qualitative assessments of phytoremediation effects on food production, income generation, livelihood strategies and farmer wellbeing are also being carried out.

Elsewhere in sub-Saharan Africa, numerous other studies have confirmed the socio-economic benefits of UPA in growing cities. Less attention, however, has focused on the health and environmental impacts of urban food production, particularly in the context of rapidly expanding urban ecological footprints across the continent. To date, there has been little research carried out in Freetown that has physically measured or evaluated these impacts, particularly in the context of wastewater reuse for irrigation at UPA sites and crop quality. This paper seeks to contribute by focusing on this under-researched area, characterising the deeply rooted connection between health and environmental impacts of UPA on water and food security. Addressing this knowledge gap is imperative, as the water used for irrigation at most major UPA sites is frequently contaminated by domestic and industrial wastewater; currently, these may be the only reliable water sources available.

During this interdisciplinary, mixed-method study, the environmental health impacts and contamination of irrigation water, arable sediment and sweet-potato leaf crop biomass were evaluated at three UPA sites in Freetown. Results show that pollutant concentrations in water, sediment and crops at the Freetown UPA study sites exceed regulatory limits, with sweet-potato crop leaves accumulating heavy metals (Ni, Cu, As, Cr, Cd, and Pb) at concentrations up to >100 times the WHO RVs for leafy greens and root vegetables. Excessively high nutrient concentrations in irrigation water were observed during the rainy season, possibly attributed to fertiliser runoff. Furthermore, microbial contamination of irrigation water was identified as a significant problem, correlating to an observed relationship between diarrhoeal infections and the absence of household water disinfection. As evidenced, surface water bodies in Freetown are one of the predominant places for liquid- and solid-waste disposal, resulting in extremely high concentrations of faecal coliform bacteria. Based on the health-related findings from the household survey, and paired with the environmental monitoring results, there is a pressing need for future studies to place more focus on the public health components of UPA. Emergent work from this study is currently investigating the feasibility of incorporating phytoremediation barriers at Freetown UPA sites. Ultimately, the study results highlight growing concerns regarding food security and public health in Freetown while also planting the seeds for potential nature-based solutions.

This research was supported by a grant from the British Academy and Royal Academy of Engineering, and executed through the University of Bath, Department of Architecture and Civil Engineering and its project partner the University of Sierra Leone, Fourah Bay College. We would like to thank the Editor of Water Supply, and two anonymous reviewers for their insightful comments on an earlier draft of this paper. We would also like to thank the late Kabba Bangura, who tragically died before this article could be published. We are grateful to Kabba for his kindness, his inspiration and his significant contribution to the fieldwork for this paper.

All relevant data are included in the paper or its Supplementary Information.

The authors declare there is no conflict.