The interaction between a putative pathogen, Aeromonas hydrophila, and drinking water system biofilm developed from dechlorinated Cincinnati tap water on polycarbonate and iron coupons was investigated. Bench scale experiments were conducted using annular biofilm reactors to mimic the hydraulic conditions inside drinking water distribution systems.
The feed tanks were emptied, then filled with fresh tap water and then dechlorinated by adding sodium thiosulfate (Na2S2O3) daily. During the biofilm development period, dechlorinated tap water was recycled through the reactor system for a 24-hr period before freshly prepared feed was added. Bulk water heterotrophic plate counts reached a stable level very quickly for the experiments with iron coupons (2 weeks), whereas a longer period of time was required for polycarbonate coupons (3-4 weeks). After spiking in high levels (e.g., 108 CFU/ml), Aeromonas hydrophila was present in biofilm samples as well as in bulk water samples for two to three weeks before decreasing below detectable levels (as measured by membrane filtration using ADA-V agar plate counts) for the polycarbonate coupons and were detectable in both, types of samples during the entire 1008 h post-spiking period for the iron coupons. Generally, biofilms grown on iron coupons contained more Aeromonas cells per unit area for a longer period of time than biofilms grown on polycarbonate coupons. Higher initial concentrations were associated with longer retention times compared with lower concentrations.
Mass balance calculations for the polycarbonate experiments revealed that growth of Aeromonas hydrophila in the reactors may have played a role in the longer retention times observed (as compared to theoretical washout rates).
Correlation between the amount of phospholipid phosphate (PLP) levels and related ADA-V colony forming unit (CFU) plate counts for Aeromonas hydrophila isolates JB-38 and ATCC 7966 compared well with published data for unspecified biofilm organisms at high concentrations, but the dilution series did not show a linear correlation, leading to speculation that PLP tests containing high numbers of CFU may be reagent-limited.
The impact of pipe material and surfactant on quantitation in the heterotrophic plate count (HPC) and phospholipid phosphate (PLP) methods was also studied. The most noticeable difference were samples into which “corrosion” was added which showed higher counts than the control versus the comparatively lower HPC counts noted for the samples into which “ferric chloride (FeCl3)” was added. Also, the biofilm samples with “Tween 80” addition showed the lowest PLP recoveries, likely due to chemical interference in the test method.