Metagenomic Analysis of Bacterial Communities in Drinking Water Distribution Systems in Mombasa County (Kenya)
Jeophita, Mwajuma June
MetadataShow full item record
Hygienic problems in drinking water distribution systems may originate from contamination by external microorganisms or growth of indigenous biomass. Among all water related disease outbreaks, 33% waterborne diseases were caused by-contaminated source water, 39% by inadequate or interrupted treatment processes and 18% by distribution systems and premise --plumbing deficiencies (Craun et 'al., 2010). The Mombasa County water distribution systems have since inception undergone a sanitary survey and there has been no information on the bacterial community composition therein . .The primary focus of this study therefore was to generate detailed information on the genome information stored within the water distribution system, microbiome. This was achieved through an analysis of the 16S rRNA phylotypes of the microbes present within the microbial community of21 water and biofilm samples collected from Mzima and Baricho water lines, which have been operational for more than 50 years, making them excellent sources of mature distribution system biofilms. Physicochemical characteristics of the water were determined by a portable Palin Photometer and Atomic Absorption Spectrophotometer; while bacterial taxonomic affiliations were analyzed using 16S rRNA based 454-FLX Titanium pyrosequencing. Water from both water lines registered pH, nitrates, phosphates and residual chlorine levels that were within the limits stipulated for drinking water by the Kenya Bureau of Standards (KEBS). All Baricho line samples had iron and lead levels way above the maximum allowable limit of 0.05 mg L-1• Nitrates, iron and temperature correlated positively with bacterial community composition and diversity in all samples as shown by Canonical correspondence analysis (Mantel test: r = 0.27, P = 0.001). Pyrosequencing yielded 27,937 sequences, which were denovo clustered into 2,294 unique operational taxonomic units (OTUs) based on their sequence similarity (3%). A total of 20 bacterial phyla and 6 candidate phyla were identified from pooled samples and were dominated by the Proteobacteria (73.2%), Firmicutes (13.4%), Bacteriodetes (5.9%) and Candidate divisions at 0.9% of the total phylotypes. Renyi diversity profiles t' demonstrated that all sampled sites regardless of source or type have larger species richness, but lower species evenness. Shannon-Wiener diversity (H) index of each sampling site ranged from 0.00 to 1.725. The highest bacterial diversity was found in Baricho water (1.725) and Mzima biofilms (1.391). Biofilms featured characteristically higher bacterial diversity, richness and abundance than bulk water. Bulk water was predominated by Nitrospirae (20.2%), Betaproteobacteria (15.9%) and Alphaproteobacteria (6.4%), while Favobacteria (8.6%), Deltaproteobcteria (3.2%), bacteria NP L. UPA2 and Candidate division OD! were characteristic of biofilms. Redundancy analysis indicated substantial comparative differences in water and biofilms bacterial community composition among the two water lines. Nitrospirae, Elusimicrobia, Cyanobacteria, Gemmatimonadetes, NP-UPA2, and Candidate Divisions TM7, OP11 and OD1 were present only at source but not at endpoints. Differences in community structure and abundance were noted between Baricho water line source and endpoint water and biofilms bacterial composition (p= 0.013). A total of 140 phylotypes of potentially pathogenic species including; Pseudomonas, Escherichia, Shigella, Aeromonas, Enterobacter and Bdelovibrio were identified. Metagenome analyses also confirmed the ubiquity of mycobacteria in drinking water distribution systems. Maintainance of the integrity of water systems, periodic monitoring and effective treatment techniques should take precedence in water delivery services to reduce the risk of contaminating the drinking water with pathogenic microorganisms.