Modeling human pollution in water bodies using somatic coliphages and bacteriophages that infect Bacteroides thetaiotaomicron strain GA17.

The ability to detect human fecal pollution in water is of great importance when assessing the associated health risks. Many microbial source tracking (MST) markers have been proposed to determine the origin of fecal pollution, but their application remains challenging. A range of factors, not yet sufficiently analyzed, may affect MST markers in the environment, such as dilution and inactivation processes. In this work, a statistical framework based on Monte Carlo simulations and non-linear regression was used to develop a classification procedure for use in MST studies. The predictive model tested uses only two parameters: somatic coliphages (SOMCPH), as an index of general fecal pollution, and human host-specific bacteriophages that infect Bacteroides thetaiotaomicron strain GA17 (GA17PH). Taking into account bacteriophage dilution and differential inactivation, the threshold concentration of SOMCPH was calculated to be around 500 PFU/100 mL for a limit of detection of 10 PFU/100 mL. However, this threshold can be lowered by increasing the analyzed volume sample, which in turn lowers the limit of detection. The resulting model is sufficiently accurate for application in practical cases involving MST and could be easily used with markers other than those tested here.