Aims: To profile fractions of active bacteria and of bacteria culturable with routine heterotrophic plate count (HPC) methods through a typical water treatment process and subsequent distribution system. In doing so, investigate how water treatment affects both bacterial abundance and diversity, and reveal the identities of active bacteria not detected by traditional HPC culture.
Methods and results: Profiling active fractions was performed by flow cytometric cell sorting of either membrane-intact (BacLight kit) or enzymatically active (carboxyfluorescein diacetate, CFDA) bacteria, followed by eubacterial 16S rDNA-directed PCR and denaturing gradient gel electrophoresis (DGGE). Water treatment significantly reduced active bacterial numbers detected by the BacLight kit and CFDA assay by 2.89 and 2.81 log respectively. Bacterial diversity was also reduced from > 20 DGGE bands in the active fractions of reservoir water to only two bands in the active fractions of finished water. These two bands represented Stenotrophomonas maltophila, initially culturable by HPC, and a Burkholderia-related species. Both species maintained measurable traits of physiological activity in distribution system bulk water but were undetected by HPC.
Conclusions: Flow cytometric cell sorting with PCR-DGGE, to assess water treatment efficacy, identified active bacteria from a variety of major phylogenetic groups undetected by routine HPC. Following treatment S. maltophila and a Burkholderia-related species retained activity and entered distribution undetected by HPC.
Significance and impact of the study: Methods used here demonstrate how water treatment operators can better monitor water treatment plant efficacy and assess distribution system instability by the detection and identification of active bacteria recalcitrant to routine HPC culture.