Characterization and source tracking of diarrheagenic bacteria contaminating fish in lake victoria, Kenya

Abstract

Lake Victoria is an important source of fish. Nonetheless, fish has been reported to be of low microbiological quality as they have been reported to be contaminated with Escherichia coli, Vibrio spp, Shigella spp and Salmonella. Nevertheless these microbial contaminants of fish in Lake Victoria have not been linked to their possible reservoirs, therefore making it difficult for regulatory agencies and fishers minimize fish contamination. Furthermore, it is not known if E. coli contaminating fish from Lake Victoria is pathogenic or not. Similarly, is not known if antibiotic resistance encountered among microbes present in Lake Victoria fish linked to clinical or environmental sources. The study aimed at characterizing and source tracking diarrheagenic bacteria contaminating fish in Lake Victoria. Fish and environmental samples from 5 locations (Sirongo, Dunga, Homa bay, Mbita town, and Luanda Konyango beaches) and human stool specimen from Kisumu East Sub-County hospital were investigated for E. coli, Salmonella spp, Shigella spp and V. cholerae. Standard microbiological procedures, API 20 E, serotyping, multiplex PCR and antibiotic susceptibility testing were used to recover and characterize the microbes. Antibiotic inhbition zones were used to discriminate isolates. E. coli was detected highest in human (92.9%) and lowest in freshly caught Nile perch and Rastriniobola argentea at 7.3% and 3.5% respectively. Shigella spp was only recovered among human, Salmonella spp was detected in soil (16.7%), sundried R. argentea (8.6%), human (7.1%). No V. cholerae was detected. High levels of total coliforms counts were recorded for water at shores (log10 3.75 ± 0.26 cfu/ml), with a significant decline from the shoreline (0m) towards offshore (150m) (p = 0.0001). Enterotoxigenic E. coli were detected among human stool. No E. coli virulence genes tested were detected among fish isolates. Overall antibiotic resistance rate of 49.7% was detected. E. coli isolates recovered from soil and R. argentea recorded the highest resistance to tetracycline (100%). Tetracycline, ampicillin, nalidixic acid and cifuroxime showed significant variation in sensitivity among different sources of E. coli (p < 0.05). Multiple Antibiotic Resistance (MAR) index of E. coli; grouped soil and fish isolates together, whereas DA with an average rate of correct classification (ARCC) of 41%, misclassified 38.5% of E. coli isolates recorvered from fish as soil isolates. The study therefore concludes that soil could be a major source of diarrheagenic bacteria contaminating Lake Victoria fish and recommeds MAR indexing of E. coli and DA as possible tools for determining sources of contamination among fish in the region.