An assessment of bacteriological quality, with special emphasis on the presence of escherichia coli, in the roof -collected rainwater from some peri-urban areas of Nairobi, Kenya

Abstract

Rainwater is g a i n i n g importance as a supplementary source of drinking water, especially in the peri-urban zones of Nairobi, where piped water is not available, or the supply is irregular. The ground water sources, boreholes and wells, may be unsuitable due to unpleasant odours and taste, or microbial contamination. Studies addressing the public health suitability of rainwater for drinking purposes are lacking in this country and the region. The present study was conducted to determine the presence of Escherichia coli [E. coli} in rainwater harvested for human consumption. This study was conducted in Ngong Division of Kajiado District, Kikuyu Division of Kiambu District and Dagoretti Division of Nairobi Province. These divisions are found in the neighbourhood of the city of Nairobi, Kenya, and w ere purposely selected . Collection of sample s was don e fro m regions where people cons urn e rain w at er stored in reservoir tanks. An area with a rich tree cover, moderate tree cover, and a pre d 0 m in an tl y res ide n t i aIr e g 10 n we res e 1e c te d for sam p li n g . Li v e s to c k k e ep in g and c r 0p far m in g was don e in the sea rea s to varying degrees. The homes selected for samplin g from were those convenientl y situ ate d alo ng th e commonly used roads in the study area. Rainwater samples were collected from water tanks. The roofs were made of either concrete roofing-tiles or galvanized corrugated i r o n sheets. The water tanks were made of galvanized iron sheets, concrete (or ferro-concrete), or plastic. Bacteriological analyses were done to determine the microbiological quality of water from the sources. The total viable count of the microorganisms was assessed using the pour plate method. The most probable number of coliforms was determined using the multiple tube fermentation technique. The presence of E. coli was determined by biochemical reactions of Indole, Methyl Red, Voges Proskauer, and fermentation of Citrate. Eighty-nine (89) water samples were collected; 29 from Dagoretti Division, 31 from Ngong Division, and 29 from Kikuyu Division. The TVC per ml of the water samples ranged between 0 and 1.02x107 in Dagoretti and Ngong while it ranged between 0 and 8.0x105 in Kikuyu. In Dagoretti, the mean TVC was 1.665x106 with a standard deviation (SD) of 2.995X106 while in Ngong it was 8.236x105 with an SD of 2.545X106 and in Kikuyu it was 7.756x104 with an SD of 1.837X105 . The most probable number (MPN) of coliforms per lOOmIs of water sample ranged between 0 and 1609 in the 3 Divisions. The means were 684 with an SD of 780 for the area, 740 and an SD of 734 for Dagoretti, 652 and an SD of 641 for Ngong, and 891 with an SD of 729 for Kikuyu. In the study area, 89% of the samples screened tested positive for general coliforms, with 40% testing positive for E. coli. Area-wise, 83% of the samples from Dagoretti tested positive for coliforms and 31 % for E. coli; 87% from Ngong were positive for coliforms and 55% for E. coli, while in Kikuyu 97% tested positive to coliforms and 34% for E. coli. An analysis of the v a r ria n c e In the TVC of water samples from the three divisions showed that there was no significant difference between Dagoretti and Ngong, but that both divisions had a significantly higher TVC than Kikuyu [p 0.033]. Of the 29 samples from Dagoretti, 24% had less than 3 coliforms per lOOmIs, 28% less than 10 coliforms, and 72% more than 10 coliforms per lOOmIs; in Ngong it was 19%,23% and 77% respectively, while in Kikuyu it was 7%, 17% and 83% respectively. Seventy-eight percent (69/89) of the samples collected contained more than 10 coliforms per lOOmIs of water, and therefore failed the Kenya Bureau of Standard's requirement for un-piped drinking water. Eighty-three percent (74/89) of the samples of the samples in the study area had more than 3 coliforms per lOOmIs of water and therefore failed the World Health Organization's standard for untreated drinking water. It was therefore concluded that rainwater from the study region should be treated before consumption.