Assessment of methods for determing the health microbiological quality of untreated surface water in the Modder River catchment

Abstract

This study compared three Escherichia coli (E. coli) detection methods to determine an optimum method to use in assessing the health-related microbiological water quality in surface water of the Middle Modder River catchment. As an outcome, the study recommends an optimum method for two types (varying pollution status) of river water. Faecal coliforms and E. coli were selected as microbiological indicator organisms. The three methods were based on colour changes to indicate the presence of the selected organism groups. Two of the methods were membrane-filtration based while the other was based on the Most Probable Number technique. The membrane-based methods used the reagent media M-FC (for faecal coliforms) and Chromocult Coliform® Agar for E. coli respectively. These were compared with the miniaturised liquid-based Most Probable Number (MPN) using the Quanti-Tray® I 2000 -Colilert 18 method for E. coli. The Colilert method turned out to be the optimum method for the water selected for this study. Higher numbers of true indicator organisms were detected by this method than by the other two methods in all the selected waters. Using the identified optimum method, the following was determined: o The E. coli numbers in the water of the Modder River (at the 90th percentile compliance limit) exceeded recommended water quality guideline limits (South African Water Quality Guidelines, the Quality of Domestic Water Supplies: Assessment Guide as well as the National Microbial Monitoring Programme) for indicator organisms in water for domestic use (drinking, food preparation, bathing and laundry) and recreation waters. The E. coli numbers did not exceed recommended water quality guideline limits for indicator organisms where the intended use was for crop irrigation. Water in the Modder River was suitable for raw water extraction to treat for use as drinking water. Abstract