Investigation into the bacterial contamination in a spring water distribution system and the application of bioremediation as treatment technology

Abstract

Spring water bottled and sold for human consumption can only be subjected to certain treatment processes such as separation from unstable constituents by decantation, filtration and aeration, ultraviolet irradiation and ozonation. A spring water distribution system in the Western Cape, South Africa was experiencing microbiological problems. The aim of the study was to investigate bacterial contamination in the spring water distribution system and the application of bioremediation as treatment technology. Sampling at various points in the spring water distribution bottling system started in February 2004 and continued until November 2004. The acceptable microbiological limits for bottled spring water clearly states that the total viable colony count should be < 100 organisms per ml of water. Analysis of samples by the heterotrophic plate count (HPC) technique indicated significantly (p < 0.05) high counts which did not conform to the microbiological limit. The heterotrophic plate counts recorded for weeks one, four, eight & 46 in the final bottled water (Site J) were 3.66 x 107 cfu/ml, 9.0 x 106cfu/ml, 2.35 x 107 cfu/ml and 5.00 x 104 cfu/ml, respectively. The total cell counts [Flow cytometry analyses (FCM)] recorded for week one, four, eight & 46 in the final bottled water (Site J) were 5.44 x 107 microorganisms/ml, 8.36 x 107 microorganisms/ml, 9.09 x 107 microorganisms/ml and 5.70 x 107 microorganisms/ml, respectively. The higher viable total cell counts(FCM) indicate that flow cytometry was able to detect cells in the water sample that enter a viable but not culturable state and that the heterotrophic plate count technique only allowed for the growth of the viable and culturable cells present in the water samples. This indicated that the HPC is not a clear indication of the actual microbial population in the water samples. It could be concluded that FCM technique was a more reliable technique for the enumeration of microbial populations in bottled water samples. Various organisms were identified by means of the Polymerase Chain Reaction (PCR) using 16S rRNA specific primers. Purified PCR amplicons were sequenced and Phylogenetic trees were constructed. Neighbour-joining phylogenetic tree analysis of the bacterial species present in the water samples was performed. The dominant bacterial isolates that were sequenced from the various water samples throughout weeks one, four, eight and 46 were Bacillus sp. and Enterobacteriaceae. The pathogenic species isolated throughout the sampling period included Escherichia sp., Pseudomonas sp., Shigella boydii, Bacillus and Staphylococcus sp. A laboratory-scale bioreactor was constructed and water samples were analysed over a period of two weeks. Water samples were analysed using FCM and Direct Acridine Orange Count (DAOC) in conjunction with epiflourescence microscopy (EM). The FCM counts ranged from 1.53 x 107 microorganisms/ml in the initial sample (Day 0) to 1.16 x 107 microorganisms/mℓ in the final sample (Day 13). The results indicated a 24% decrease in the microbial numbers however, it was still above the limit of < 100 organisms/ml as set out by the South African Standards of Bottled Water, (2003). The total cell counts obtained by the DAOC method ranged from 1.43 x 106 microorganisms/ml to 9.54 x 105 microorganism/ml on day 13 (final). The results indicated a 33% decrease in microbial numbers. The total cell counts analysed by flow cytometry fluctuated throughout the sampling period. The total cell counts obtained from the DAOC method were lower in all the water samples when compared to the total counts obtained by flow cytometric analyses. Even though the FCM counts fluctuated throughout the sampling period, results clearly show that the FCM method yielded more accurate data for total cell counts than the DAOC method. Due to external environmental conditions such as changes in the weather conditions the results fluctuated and the final results clearly indicated that further studies are required to optimise the bioreactor system for its application in the spring water industry.