Search

Despite being a basic human right, limited access to clean water is still a major concern in developing countries lacking adequate sanitary infrastructure. A significant proportion of the global population directly depends on surface water resources which are often contaminated with fecal matter. The presence of fecal contamination in waterbodies is often detected using fecal indicator bacteria like Escherichia coli. According to 2016 UNEP report, about one third to one half of Asian rivers are estimated to be severely polluted, with monthly in-stream concentrations of fecal coliform bacteria exceeding 1000 cfu.100 mL-1. Although various studies on small tropical catchments have improved our understanding of E. coli behavior in a tropical context, little information exists on the underlying mechanisms at large watershed scales during dry and wet seasons. Our study focuses on Mekong River and its main tributaries in Laos, an area that has witnessed rapid changes in land use and deterioration of water quality over the last three decades. We aim (1) to examine the seasonality of E. coli concentrations in stream waters, and (2) to identify the main factors controlling E. coli in-stream concentration, such as land use, hydrometeorology, and suspended sediment concentrations, through field monitoring of a range of catchments across Laos. To this end, we used two different sets of field data monitoring at multiple temporal and spatial scales. First, a total of 18 catchment outlets located between 15°N and 20°N, were sampled twice in 2016, during both dry and rainy seasons, covering a broad range of catchment sizes (240 - 25946 km²), as well as geographical and topographical features. Second, three northern rivers, Nam Ou, Nam Suang, and Mekong River, have been sampled every 10 days since July 2017. Our results shed the light on contamination over the year in all three catchments (100-100000 MPN.100 mL-1), with higher E. coli concentrations during the rainy season, associated with higher water levels, and higher concentrations of total suspended sediment (TSS) in streams. Partial Least Square (PLS) regression showed a strong positive correlation between E. coli concentrations and the percentage of unstocked forests area. Unstocked forests are exposed to erosion processes resulting in high concentrations of suspended sediment and particle-attached E. coli in-stream concentrations. In contrast, catchments with larger protected and naturally regenerated forest and grassland areas were associated with lower E. coli and TSS concentrations. These analyses highlight the importance of adequate land management in tropical context to reduce soil loss and water quality degradation. Furthermore, our results reveal the importance of improving our understanding of fate and transport of fecal contamination through field monitoring at various spatial and temporal scales, in order to assess the risk to public health, and the impact on ecosystem services, such as contaminant retention.