Exploring carbon dynamics in a slow sand filter using stable isotopes.

Slow sand filtration (SSF) is one of the oldest biofiltration methods for reducing pathogens and organic matter (OM) in water. Due to its efficiency, affordability, and operational simplicity, SSF remains a widely used approach for producing biologically stable drinking water. Although biological activity plays a role in the removal of OM during SSF, its contribution is poorly constrained. Here, we explored the utility of stable isotopes for investigating this role quantitatively on the scale of an operational filter. First, by combining measurements of concentrations and natural isotopic composition in relevant carbon pools (dissolved and solid, organic and inorganic), we found evidence for OM removal through both retention and subsequent mineralization. However, their relative contributions could not be constrained due to insufficient precision and continuity of available data and incomplete knowledge about the relevant isotope fractionation factors. In the other approach, we therefore used laboratory incubations of SSF cores with C-labeled glucose over 14 days and found rapid removal of the tracer by the biological community, exceeding the assimilable organic carbon loading rate of the operational filter by 18 times. The glucose removal was not limited to the upper part of the sand column, the schmutzdecke, but occurred throughout the entire sand column. Furthermore, the removal was dominated by bacterial uptake over mineralization, with a substantial part likely retained as carbon reserves. The residence time of the tracer exceeded the duration of the experiment, hampering our ability to estimate the rate of OM mineralization. Analysis of the meiofauna indicated that grazing and/or predation constitutes only a minor sink for the bacterial biomass in the studied filter. Overall, this study illustrates the potential of stable isotopes for studying biological processes in SSF systems, including OM removal under diverse conditions, maturation of new or recently cleaned filters, or interactions within the endogenous biological community. To fully utilize this potential, future work should employ isotope labeling experiments with a longer duration, and consider more systematic and precise monitoring of the concentrations and isotopic composition in the relevant carbon pools.