Transport of E. coli D21g with runoff water under different solution chemistry conditions and surface slopes

- Exchange with the sand increased with the runoff chamber slope.
- Bacteria transport and release with runoff water depended on the chemistry.
- Cell retention increased with ionic strength and slope.
- The transient storage model accurately described the transport data.
- Improved understanding of exchange was obtained using a fully coupled model.

SummaryTracer and indicator microbe runoff experiments were conducted to investigate the influence of solution chemistry on the transport, retention, and release of Escherichia coli D21g. Experiments were conducted in a chamber (2.25 m long, 0.15 m wide, and 0.16 m high) packed with ultrapure quartz sand (to a depth of 0.10 m) that was placed on a metal frame at slopes of 5.6%, 8.6%, or 11.8%. Runoff studies were initiated by adding a step pulse of salt tracer or D21g suspension at a steady flow rate to the top side of the chamber and then monitoring the runoff effluent concentrations. The runoff breakthrough curves (BTCs) were asymmetric and exhibited significant amounts of concentration tailing. The peak concentration levels were lower and the concentration tailing was higher with increasing chamber slope because of greater amounts of exchange with the sand and/or extents of physical nonequilibrium (e.g., water flow in rills and incomplete mixing) in the runoff layer. Lower amounts of tailing in the runoff BTC and enhanced D21g retention in the sand occurred when the solution ionic strength (IS) was 100 mM NaCl compared with 1 mM NaCl, due to compression of the double layer thickness which eliminated the energy barrier to attachment. Retained cells were slowly released to the runoff water when the IS of the runoff water was reduced to deionized water. The amount and rate of cell release was greatest at the highest chamber slope, which controlled the amount of exchange with the sand and/or the extent of physical nonequilibrium in the runoff layer, and the amount of retained cells. The observed runoff BTCs were well described using a transient storage model, but fitted parameters were not always physically realistic. A model that accounted for the full coupling between flow and transport in the runoff and sand layers provided useful information on exchange processes at the sand surface, but did not accurately describe the runoff BTCs which were influenced by physical nonequilibrium in the runoff layer.