Exploration of how UV disinfection affects the microbial community in drinking water

Abstract: UV irradiation is getting increasingly popular as a disinfection step in drinking water production. Ringsjöverket, a drinking water treatment plant in Sweden installed a full-scale UV aggregate in December 2016. However, the effect of UV irradiation on drinking water has mainly been tested on specific pathogens in labscale environment. Heterotrophic plate counts are often used when analyzing drinking water despite the fact that the results may not always relate to the total bacteria population or the existence of harmful pathogens. This study investigated the effect of UV on the whole bacterial community with heterotrophic plate counts (HPCs), flow cytometry (FCM) and qPCR. Untreated water directly before Ringsjöverket UV aggregate, water treated with Ringsjöverket ́s UV aggregate and water treated with a labscale UV aggregate was analyzed. Water was stored in light-impermeable 20 l water containers at 7 °C and 22 °C. The water was then analyzed as a function of days after UV irradiation up to 10 days. HPC resulted in that the UV treated samples showed less diversity in terms of morphology of CFUs, but also had more CFUs. This could imply that UV has a selective effect. FCM analysis of intact cells pointed to that the UV treated samples did not grow back, but also did not die off completely. This can be compared with an untreated reference where bacteria count increased from initial levels. The results from the intact cell analysis implies that UV inhibits the growth of bacteria but does not necessarily have a direct killing effect. Flow cytometric analysis pointed towards a decrease in the fraction of cells with high nucleic acid content compared to untreated reference. At higher temperature the decrease was followed by an increase a couple of days later. This implies that UV initially changes the bacteria community composition but that the effect is not permanent due to repair mechanisms or that some bacteria does not get affected by UV and then start growing. The instant effects of UV were not detected by HPC or FCM but seen with qPCR targeting the 16S rRNA gene. This method showed that untreated samples had more DNA that could be amplified, and that Ringsjöverket UV seemed to result in less DNA damage than the labscale aggregate. Further studies could be directed at investigating different dyes for flow cytometric analysis, such as dyes that targets bacteria at different rates depending on metabolic activity. Also, detection of UV damage through qPCR could be improved through using longer amplicon length. To get even further insight in how the taxonomic composition of bacteria is changed by UV, the amplicons from qPCR can be sequenced.