Shedding light on the total and active core microbiomes in slow sand filters for drinking water production

Abstract

Slow sand filters (SSF) are widely used in the production of drinking water as a last barrier in the removal of pathogens. This removal mainly depends on the ‘Schmutzdecke’, a biofilm-like layer on the surface of the sand bed. Most previous studies focused on the total community as revealed by DNA analysis rather than on the active community, which may lead to an incorrect understanding of the SSF ecology. In the current study, we determined and compared the DNA- (total) and RNA-displayed (active) communities in the Schmutzdecke layer from 10 full-scale slow sand filters and further explored the SSF core microbiome in terms of both presence (DNA) and activity (RNA). Discrepancies were observed between the total and the active community, although there was a consistent grouping in the PCoA analysis. The DNA-displayed community may be somewhat inflated, while the RNA-displayed community could reveal low abundance (or rare) but active community members. The overall results imply that both DNA (presence) and RNA (activity) data should be considered to prevent the underestimation of organisms of functional importance but lower abundance. Microbial communities of studied mature Schmutzdecke were shaped by the influent water. Nevertheless, a core microbiome was shared by the mature Schmutzdeckes from independent filters, representing the dominant and consistent microbial community composition in slow sand filters. In the DNA samples, a total of 33 VSC families (‘very strict core’, with a relative abundance >0.1% and 100% prevalence) were observed across all filters. Among the RNA samples, there were 18 VSC families, including 16 families that overlapped with the DNA VSC families and 2 unique RNA VSC families. The core microbial community structure was influenced by the operational parameters, including the Schmutzdecke age and the sand size, and was less influenced by water flow. In addition, indicator organisms (‘biomarkers’) for the Schmutzdecke age, which show the longest duration that SSF can maintain a good operation, were observed in our study. The abundant presence of bacteria belonging to bacteriap25 and Caldilineaceae was associated with older Schmutzdeckes, revealing longer periods of stable operation performance of the filter, while the high abundance of bacteria belonging to Bdellovibrionaceae and Bryobacteraceae related to short periods of stable operation performance.