An interdisciplinary toolbox to track the sources of fecal contaminations in water

International audienceMicrobiological contamination of water is of particular concern in sensitive areas such as bathing or shellfish waters. The presence of pathogen microorganisms can originate from animal waste (release from cowpats during rainfall or in runoff during a rainfall event after manure application) or from discharge of effluent from wastewater treatment plants (WWTP). Fecal contaminations of inland and coastal waters induce risks to human health and economic losses. In order to improve water management, it is necessary to identify the sources of contamination, which implies the development of specific markers. In order to be considered as a valuable host-specific marker, one must (1) be source specific, (2) occur in high concentration in polluting matrices, (3) exhibit extra-intestinal persistence similar to fecal indicator bacteria (FIB) and (4) not grow out of the host. However, up to day no single marker has fulfilled all those criteria. Thus, it has been suggested to use a combination of markers in order to generate more reliable data. The aim of this study was to develop a microbial source tracking (MST) toolbox to discriminate between three main sources of fecal contamination (porcine, bovine and human). Chemical markers (caffeine and steroid fingerprint) and microbiological markers (Bacteroidales, Lactobacillus amylovorus and F-specific RNA bacteriophage genogroups) were selected for their host specificity. The development of this MST toolbox was composed of four steps, from the molecular scale to the watershed scale. At the molecular scale, the specificity and the concentration of those markers were studied in cattle and pig manures and in waste water treatment plant (WWTP) effluents and influents. At the microcosm scale, the transfer of bovine and porcine specific markers was investigated by rainfall simulations on agricultural plots amended with cattle or pig manure. Moreover, the relative persistence of FIB and human, porcine and bovine specific markers was investigated in freshwater and seawater microcosms inoculated with WWTP influent and pig manure. Finally, the aforementioned MST toolbox has been validated at the catchment scale by analysing rivers and a rural watershed located in Brittany (France), over a year. Under the conditions of the microcosm experiment (maintained aerobically in the dark at 18 C), bacterial markers persisted in waters contaminated with pig manure and WWTP effluent at least 21 and 6 days, respectively. The steroid fingerprint indicated a human or a porcine contamination during 13 days. The caffeine was also detected at least 13 days. Results of the analyses carried out in the watershed highlighted that human contamination was regularly detected whereas bovine and porcine contamination were intermittent. It is noteworthy that the ruminant markers were positively correlated with the rainfall contrary to the human markers. The latter were positively correlated to the fecal indicators. In conclusion, the combined use of chemical and microbiological markers can differentiate three main sources of pollution which may alter the quality of water in Brittany