Abstract
Legionella spp. are common in various natural and man-made aquatic environments. Recreational hot spring is frequently reported as an infection hotspot because of various factors such as temperature and humidity. Although polymerase chain reaction (PCR) had been used for detecting Legionella, several inhibitors such as humic substances, calcium, and melanin in the recreational spring water may interfere with the reaction thus resulting in risk underestimation. The purpose of this study was to compare the efficiencies of conventional and Taqman quantitative PCR (qPCR) on detecting Legionella pneumophila in spring facilities and in receiving water. In the results, Taqman PCR had much better efficiency on specifying the pathogen in both river and spring samples. L. pneumophila was detected in all of the 27 river water samples and 45 of the 48 hot spring water samples. The estimated L. pneumophela concentrations ranged between 1.0 × 10(2) and 3.3 × 10(5) cells/l in river water and 72.1-5.7 × 10(6) cells/l in hot spring water. Total coliforms and turbidity were significantly correlated with concentrations of L. pneumophila in positive water samples. Significant difference was also found in water temperature between the presence/absence of L. pneumophila. Our results suggest that conventional PCR may be not enough for detecting L. pneumophila particularly in the aquatic environments full of reaction inhibitors.