Water quality and public health in northern Sudan: a study of rural and peri-urban communities

Transcription

1 Journal of Applied Microbiology 1999, 87, Water quality and public health in northern Sudan: a study of rural and peri-urban communities H.A. Musa, P. Shears 1, S. Kafi and S.K. Elsabag Department of Medical Microbiology and Parasitology, Faculty of Medicine, University of Khartoum, Khartoum, Sudan and 1 Centre for Tropical Medical Microbiology, Department of Medical Microbiology and School of Tropical Medicine, University of Liverpool, Liverpool, UK 7153/04/99: received 6 April 1999, revised 16 June 1999 and accepted 28 June 1999 H.A. MUSA, P. SHEARS, S. KAFI AND S.K. ELSABAG Access to adequate supplies of good quality drinking water continues to be limited among many rural and periurban communities in Africa, despite several decades of water improvement programmes. The present study investigated water quality at the source and point of consumption among rural and peri-urban communities in northern Sudan. Faecal coliform counts were determined by the membrane filtration technique and geometric mean counts compared in different seasons and among the different communities. Among nomadic pastoralists and riverine villages, both water sources and water stored for consumption had faecal coliform counts grossly in excess of WHO standards, with higher counts at the end of the rainy season. In the peri-urban community on the outskirts of Omdurman, while water quality from the distribution system had faecal coliform counts generally below 10 dl 1, after storage, water was of considerably lower quality, with faecal coliform counts up to 1000 d1 1. The highest counts again occurred in the rainy season. Rates of diarrhoeal disease for Khartoum province were also greatest towards the end of the rainy season. The study has shown that poor quality water continues to be a major risk factor for public health in these communities. INTRODUCTION The provision of drinking water of adequate quality and quantity remains a major public health need in many African countries, where diarrhoeal diseases continue to cause extensive morbidity and mortality. A number of studies have investigated water supply and public health in sub-saharan Africa (Cairncross and Cliff 1987; Shier et al. 1966), but relatively few data are available from the more sparsely populated semiarid Savannah regions. The Sudan is the largest country in the African continent, covering an area of 2 5 million km 2 and extending from a latitude of 5 N in the humid, tropical south to 23 N in the arid desert region of the north. In northern Sudan, extending from latitude 12 N, apart from Correspondence to: P. Shears, Centre for Tropical Medical Microbiology, Department of Medical Microbiology and School of Tropical Medicine, University of Liverpool, PO Box 147, Liverpool L69 3BX, UK ( shears@liv.ac.uk). the urban dwellers of Khartoum and Omdurman, the majority of the population are rural. North of Khartoum this population comprises two groups, the sedentary cultivators living in the villages along the river Nile and the nomadic herdsmen living in the vast desert areas, who follow traditional migration patterns determined by rainfall and available grazing. In such arid areas, limited access to water of adequate quantity and quality is a major risk factor in environmental health (Bannaga and Pickford 1978). In recent years, there has been a large movement of displaced groups into the peripheral areas of the cities of Omdurman and Khartoum. These communities live in localities characterized by crowding, poor housing and inadequate water and sanitation, although great attempts are being made by the government to improve the infrastructure. The present study was undertaken to investigate water quality in these three environments: a riverine village, a nomadic area and crowded peri-urban settlements The Society for Applied Microbiology

2 WATER QUALITY IN SUDAN 677 MATERIALS AND METHODS Study area and population Figure 1 shows the study area in northern Sudan. Wadramli is a village with a population of approximately 5000 situated on the bank of the river Nile, 70 km north of Khartoum. The majority of the villagers are cultivators, with fields irrigated by water drawn from the Nile and distributed by means of earthen canals. Water is obtained and distributed in the village in two ways. In the improved area, water is pumped from the Nile to an overhead storage tank. Connected to this are both a small chlorination unit and a slowsand filter bed, but neither of these had been functioning for some years before the study period. From the storage tank a piped distribution system supplies individual households by taps located in the open compound attached to the house. Water from the taps is stored in traditional earthen jars (zir). In the unimproved area, water is taken directly from the irrigation channels and again stored in zirs prior to consumption. El-Rawian is a major watering area for nomadic herdsmen living in the desert area between the Nile and the Atbara rivers. The water source is a rain-fed natural surface reservoir (hafir). The hafir contains water from the end of the rainy season until February or March, and is the source of drinking water for both livestock and humans. For human use, water is collected in a traditional vessel made from animal hide, (girba). The nomadic communities live in small family camps in the desert, mostly concentrated between 10 and 50 km from the hafir. Towards the end of the dry season, as the hafir dries, the nomadic groups move towards the rivers or irrigated areas and girbas are filled directly from canals. Water holes dug in the dry bed of the hafir may continue to be used for watering goats. The urban population was studied at three localities, Ombada, Marzoog and Mayo. Each was a peri-urban area with recently settled communities. The urban community was studied throughout 1995 and the rural areas in October 1995 and May Water sampling and microbiology Water samples were collected in sterile 50-ml containers from each of the sampling locations. Samples were kept in cold boxes (temperature 8 10 C) during transportation back to the laboratory of the Department of Medical Microbiology, University of Khartoum. Samples were either processed on the day of collection or kept at 4 C and processed the following day. Faecal coliforms were determined by the membrane filtration method using standard techniques (Anon. 1994) and the faecal coliform count per 100 ml (FC dl 1 ) calculated. Samples from the Nile, canals and hafir were diluted 1:5 with sterile distilled water before filtration. Samples from zirs, girbas and taps were diluted 1:2 before filtration. Faecal coliform counts per 100 ml were determined for each sample. Statistical analysis For each sampling site the geometric mean (GM) of the FC dl 1 was determined. Faecal coliform levels at different times sites were compared using the Wilcoxon rank sum test. RESULTS Environmental and climatic data The months of October and May represent distinct periods in the seasonality of northern Sudan. Figure 2 shows the mean monthly temperature and precipitation for the study area and flow data for the river Nile. October is the end of the rainy season and the beginning of the dry season which may last for 8 or 9 months. In the desert area, surface water hafirs are at their fullest, assuming that there has been sufficient rain, and are a major concentrating area for nomadic communities and their livestock. The hafir at this time had a surface area of approximately m 2 and a depth up to 2m. The month of May is the hottest part of the dry season and daytime shade temperatures may reach 48 C and night temperatures may not fall below 35 C. Flow in the river Nile is at its lowest at this time. Within the desert area the hafir was dry except for a hand-dug water hole in its bed used for goats and the few children that herded them. Water quality data Table 1 shows the GMs and ranges of the faecal coliform counts in the village and nomadic areas. For all sampling points, the faecal coliform counts were significantly higher (P ³ 0 05) at the end of the rainy season in October than in the dry season in May. At both seasons in the village area supplied by the pipe distribution system, faecal coliform counts were significantly lower (P ³ 0 05) in the storage vessels (zirs) than in the source or distribution system. In those areas of the village where water was taken from the canals, the coliform counts were reduced in the zirs compared with the canals, but the differences were not significant. In both seasons the water quality in the zirs from the canals was significantly worse than in the pipe-fed zirs (P ³ 0 05). In the nomadic area, the hafir was only available as a water

3 678 H.A. MUSA ET AL. Fig. 1 Location of project areas in Sudan source for a few months after the rains. By May it was reduced to a small water hole in its dried-up bed, used for watering occasional flocks of goats. In October, when the hafir was being used for human consumption, the faecal coliform counts were lower in the girbas (GM 676 dl 1 ) than in the hafir (GM 1288 dl 1 ), but the number of samples was insufficient to show a significant difference. In both the village and the nomadic areas, the faecal coliform counts of the water in the storage vessels were in excess of WHO (1985) guidelines for non-improved drinking supplies in tropical areas. Table 2 shows the water quality data for the three urban areas. The areas of Ombada and Marzoog were supplied by water standpipes connected to the treated municipal supply. From the standpipes, water was either taken directly into storage zirs or taken to the houses in barrels by water-sellers. In Mayo, water was taken from a tube-well. The municipal supply was of reasonable quality at most times of collection, with geometric means generally below 10 cfu dl 1 and no marked seasonal variation. The quality of water in the zirs was quite different. In all cases, water quality was significantly worse (P ³ 0 05) in the zirs than the supply water, suggesting contamination of the zirs in the household. There was also a seasonal variation of the quality of water in the zirs, with faecal coliforms being lowest in the winter and highest during the rainy season, the differences between counts in the winter and the rainy season being significant (P ³ 0 05) in each case. As was found in the rural area, the water quality in the zirs was lower than acceptable WHO standards (WHO 1985). Diarrhoeal prevalence No data were collected from study households on cases of diarrhoea, but data obtained for Khartoum province, in which the sampling sites were located, are shown in Fig. 3. The peak incidence of diarrhoea cases occurred at the end of the

4 WATER QUALITY IN SUDAN [Temperature ( C) (1400 h)] Rainfall mm Nile flow (% max velocity) J F M A M J J A S O N D 0 Month Fig. 2 Climatic data for Sudan (source Meteorological Department, Khartoum). Ž, Daily temperature; ;, monthly rainfall; ž, Nile flow rainy season, when faecal coliform counts at all sites tested were highest. These data were obtained from hospital and health centre records, and so do not represent all cases of diarrhoea in the community. While there is always some inward and outward migration of the population in the province, no major population changes occurred, suggesting that the data do represent a true seasonal trend in diarrhoea cases. DISCUSSION The provision of drinking water of adequate quality and quantity has been a major policy objective of the World Health Organisation since the Water and Sanitation decade (Cairncross 1987). Despite major efforts by international and national development programmes to achieve these aims, many communities in the tropics continue to use unimproved surface water sources for drinking purposes. Northern Sudan is typical of many of the semi-arid regions of Africa, with both sedentary and nomadic communities dependent on water supplies limited by quality and/or quantity and by migration to the poor peri-urban areas of larger towns and cities. In each of the areas of the present study, water that was consumed contained faecal coliforms considerably in excess of the WHO recommended guidelines. Such high faecal coliform counts for potable water have been found in several earlier studies in the tropics. In a review of water samples from rivers, ponds and wells in several tropical countries, Feachem (1980) reported FC dl 1 greater than at some sites with many sources in the region with FC dl 1, similar to those found in this study. Seasonal variations of water quality have been investigated in several tropical areas. Studies in Sierra Leone (Wright 1986), Gambia (Barrel and Rowland 1987) and Nigeria (Blum et al. 1987) have all shown the highest faecal coliform counts in surface water sources after the start of the rainy season. The association of high intensity rains causing run-off from faecally polluted dry soils is given as the explanation of these findings. Such relationships may vary where rainfall patterns and environment differ. In Kenya, Muhammed and Morrison (1975) showed no significant difference in faecal coliform counts in surface water sources between the dry and wet seasons. In the present study, we have investigated the seasonal variations in both water sources and water stored for consumption, in the different environments of a riverine village, a desert community and a peri-urban settlement. While in each location faecal coliform counts were highest in

5 680 H.A. MUSA ET AL. Fig. 3 Monthly diarrhoea rates for Khartoum Province, 1996 May October Site n GM Range n GM Range Village area Nile Pipe-supplied taps Zirs from taps Canal Zirs from canal Table 1 Faecal coliform counts (geometric mean (GM) and ranges) from the village and nomadic areas (counts dl 1 ) Nomadic area Hafir Girbas from hafir Girbas from canal n, Number of samples taken. the rainy season (except for the sources in the urban area) they represent different ecological settings. In the riverine villages, the higher faecal coliform counts in the Nile and feeder canals in the rainy season are probably due to the flushing effect of the rains on surface contamination. The high faecal coliform counts in the Nile result not

6 WATER QUALITY IN SUDAN 681 Table 2 Faecal coliform counts (geometric mean (GM) and ranges) from the urban areas (counts dl 1 ) Summer (March June) Rains (July October) Winter (November February) Area Sample n GM Range n GM Range n GM Range Ombada Tap Zir Marzooq Tap Zir Mayo Well Zir just from the local environment and rainfall patterns, but will also be influenced by the wider catchment and tributaries. Hydrological studies of the Nile (Rai and Hill 1978) have also demonstrated highest faecal coliform counts in the rainy season. The high faecal coliform counts in the village storage zirs in the rainy season are a reflection of the higher faecal coliform counts in their sources at this time. Generally, storage in zirs led to reductions in faecal coliform counts although, in a few households, zir faecal coliform counts were higher than those in the tap supply. If water is not further contaminated within the household, storage may reduce faecal coliform counts as bacterial growth decreases, a result found in studies in Nigeria (Tomkins et al. 1978). In the nomadic community, the hafir is the major source of water for both humans and livestock from the rainy season until February or March. The high faecal coliform counts seen in October may be caused primarily by animal rather than human faecal contamination. This may not, however, reduce the usefulness of the faecal coliform count as an indicator of risk of human disease. Outbreaks of diarrhoea in humans, caused by Escherichia coli 0157, have been reported in southern Africa resulting from animal faecal contamination of surface waters used by both humans and domestic livestock (Isaacson et al. 1993). Faecal coliform counts were lower in the storage girbas than in the hafir, suggesting a decline in bacterial survival with storage. In the peri-urban areas, the faecal coliform counts of the water sources (municipal pipe supply and wells) were less than in the rural areas, with GM faecal coliform counts less than 30 dl 1 and with no consistent seasonal pattern. The wells in Mayo were protected wells, with little risk of surface contamination. In each location, faecal coliform counts in the storage zirs were significantly higher than in the sources, suggesting contamination after collection. Several studies in the tropics have demonstrated contamination of water storage vessels within the household (El-Atar et al. 1982, Egypt; Shears et al. 1995, Bangladesh). The zirs are similar to the storage vessels in the above studies, with a wide opening at the top from which water is taken, often by hand-held containers. In poor communities, where soap for hand cleaning may not be available and where, because of crowding and inadequate sanitation, the household environment is unhygienic, there will be considerable scope for storage water contamination. In each case faecal coliform counts in the storage zirs were highest in the rainy season, although source faecal coliform counts were not elevated, suggesting a worsening environment within and around the household. Contamination of stored water by faecal contamination within the household has been shown to be an important route for the transmission of enteric pathogens (Deb et al. 1982). The relationship between faecal coliform counts and public health in the tropics, in particular with rates of diarrhoeal disease, is complex because of the many factors involved in the epidemiology of enteric pathogens (Kolsky 1993). Some studies have shown that improvements in water quality alone may not reduce the incidence of diarrhoeal diseases (Levine et al. 1976). Other studies have shown an association of diarrhoeal rates with deteriorating water quality (Young and Briscoe 1987, Malawi; Henry and Rahim 1990, Bangladesh), although other hygiene factors are likely to have been involved. The diarrhoea data for the province suggest a temporal association between diarrhoea incidence and water quality, although many other factors would be implicated in the less hygienic environment that is present during the rainy season. Interpretation of water quality data in the tropics is further complicated by the difficulties of distinguishing human from animal contamination (Mara and Oragui 1985) and the true significance of faecal or thermotolerant coliforms as indicators of non-environmental contamination in areas with ambient temperatures in excess of 44 C (Lavoie 1983). This study has shown that the communities investigated, which are representative of many in the semi-arid and sahel region of Africa, are still exposed to water supplies that pose a major risk for the transmission of water-borne diseases. This is despite both the extensive scientific understanding of

7 682 H.A. MUSA ET AL. water and health in the tropics, that was well described almost two decades ago, and the priority given to improvements in water supply (although less for sanitation) by the World Health Organisation and other development programmes. Improvements in both water supply quality and availability and the reduction of contamination within households will be necessary for the long-term improvement in public health in these communities. ACKNOWLEDGEMENTS The authors wish to thank the British Council, Khartoum, the Gordon Memorial College Trust Fund and Oxfam (UK) for their support to the field and laboratory work of this study. The authors also acknowledge colleagues in the Department of Medical Microbiology and Parasitology, University of Khartoum, for their participation in the study and Mrs N. Lowe for preparing reagents and equipment in Liverpool. REFERENCES Anon. (1994) The Microbiology of Water Part 1: Drinking Water. Report on Public Health and Medical Subjects, no. 71. London: HMSO. Bannaga, S.E.I. and Pickford, J. (1978) Water-health relationships in Sudan. Effluent and Water Treatment Journal 18, Barrel, R.A.E. and Rowland, M.G.M. (1987) The relationship between rainfall and well water pollution in a West African (Gambian) village. Journal of Hygiene 83, Blum, D., Huttly, S.R.A., Okoro, J.I., Akujobi, C., Kirwood, B.R. and Feacham, R.G. (1987) The bacteriological quality of traditional water sources in north-eastern Imo State, Nigeria. Epidemiology and Infection 99, Cairncross, S. (1987) Water supply and sanitation: an agenda for research. Journal of Tropical Medicine and Hygiene 92, Cairncross, S. and Cliff, J.L. (1987) Water use and health in Mireda, Mozambique. Transactions of the Royal Society of Tropical Medicine and Hygiene 81, Deb, B.C., Sircar, B.K. and Sengupta, P.G. (1982) Intra-familial transmission of Vibrio cholerae in Calcutta slums. Indian Journal of Medical Research 76, El-Atar, L., Gawad, A.A., Khairy, A.E.M. and el-sebaie, O. (1982) The sanitary condition of rural drinking water in a Nile Delta village. Journal of Hygiene 88, Feachem, R.G. (1980) Bacterial standards for drinking water quality in developing countries. Lancet 2, Henry, F.J. and Rahim, Z. (1990) Transmission of diarrhoea in two crowded areas with different sanitary facilities in Dhaka, Bangladesh. Journal of Tropical Medicine and Hygiene 93, Isaacson, M., Canter, P.H., Effler, P., Arntzen, L., Bomans, P. and Heenan, R. (1993) Haemorrhagic colitis epidemic in Africa. Lancet 1, 961. Kolsky, P.J. (1993) Water, sanitation and diarrhoea; the limits of understanding. Transactions of the Royal Society of Tropical Medicine and Hygiene 87 (Suppl. 3), Lavoie, M.C. (1983) Identification of strains isolated as total and faecal coliforms and comparisons of both groups as indicators of faecal pollution in tropical climates. Canadian Journal of Microbiology 29, Levine, R.J., Khan, M.R., D Souza. S. and Nalin, D.R. (1976) Failure of sanitary wells to protect against cholera and other diarrhoeas in Bangladesh. Lancet 2, Mara, D.D. and Oragui, J. (1985) Bacteriological methods for distinguishing between human and animal faecal pollution of water; results of fieldwork in Nigeria and Zimbabwe. Bulletin of the World Health Organisation 63, Muhammed, S.I. and Morrison, S.M. (1975) Water quality in Kiambu District, Kenya. East African Medical Journal 52, Rai, H. and Hill, G. (1978) Bacteriological studies on Amazonas, Mississippi and Nile Waters. Archives Hydrobiologica 81, Shears, P., Hussein, M.A., Chowdhury, A.H. and Mamun, K.Z. (1995) Water sources and environmental transmission of multiply resistant enteric bacteria in rural Bangladesh. Annals of Tropical Medicine and Parasitology 89, Shier, R.P., Dollimore, N., Ross, D.A., Binka, F.N., Quigley, M. and Smith, P.C. (1966) Drinking water source, mortality and diarrhoea morbidity among young children in northern Ghana. Tropical Medicine and International Health 1, Tomkins, A.M., Drasar, B.S., Bradley, A.K. and Williamson, W.A. (1978) Water supply and nutritional status in rural northern Nigeria. Transaction of the Royal Society of Tropical Medicine and Hygiene 72, WHO (1985) Guidelines for Drinking Water Quality, Vol. 3. Geneva: WHO. Wright, R.C. (1986) The seasonality of the bacterial quality of drinking water in Sierra Leone. Journal of Hygiene 96, Young, B. and Briscoe, J. (1987) A case-control study of the effect of environmental sanitation on diarrhoea morbidity in Malawi. Journal of Epidemiology and Community Health 42,