Overall, all 18 sample sites tested positive for enteric virus at least three times during the study period. There were 71 total viral detections during the rainy season, and 45 total detections during the dry season. Among these 18 sites, S and Isl and, a site directly next to a wastewater treatment plant and its outfall, tested positive most frequently during the study period, with a total detections of 10 times. Hale'iwa's Ali'i Beach Park on the North Shore tested positive for enteric viruses for only 3 times, representing the least number of viral detection during the study period(Table 1). Norovirus genogroup Ⅰ was detected the most often overall, followed by enteroviruses, which is fairly consistent with previous studies(Table 2)(Tong et al., 2011; Viau et al., 2011; Connell et al., 2012).
Table 1. Seasonal detection of enteric virus from selected environmental water sites around Oahu Island (by Month).
Table 2. Summary of detection of enteric viruses by types from the 18 selected recreational waters.
There was a significant difference in overall detection between the rainy and dry seasons, with 71 total viral detections in the rainy season and 45 in the dry season(P = 0.0052, Figure 2). Additionally, there was significantly more viral detection in the rainy season than the dry season based on water type defined as freshwater, recreational, near freshwater output, or near sewage treatment outfalls(P = 0.008, Figure 3). There were no statistically significant variations in detection based on virus type(P = 0.08). The majority of viruses detected did not show significant variation with season; however, norovirus genogroup Ⅱ showed an almost statistically significant variation between rainy and dry season, with more detections in the rainy season compared to the dry season(P = 0.053).
Figure 2. Total viral detections in the rainy season versus dry season. Presented as absolute number of detections of all viral types detected during rainy and dry season by site. * P = 0.005 for average total detections in rainy season as compared to dry season.
Figure 3. Percent of total possible detections in rainy season versus dry season, by water type. Percent of total possible detections calculated as total number of detections divided by number of site types and times number of possible viral types. * P = 0.008 for average total detections in rainy season compared to dry season, stratified by water type.
Selected positive samples were sequenced to confirm accurately viral detection. Several viral subtypes were found, and sequenced samples were positively identified as the virus of interest(Supplementary Table S1). In addition, a control test revealed that all the samples tested positive for E. coli, indicating successful extraction of sample nucleic acids.
Of the seventeen positive norovirus genogroup Ⅱ samples tested using qRT-PCR, four samples were positively quantified by qRT-PCR basing on the sensitivity limitation of the method. Copy numbers could only be detected down to approximately 200 copies per reaction, whereas the detection limit for PCR was likely 10-7, requiring only 0.1-10 pg of viral cDNA to obtain a positive detection(Figure 4, 5). All samples collected during the February 2013 sampling time were positively quantified and showed a range from a low of 229 copies per 25-μL reaction to a high of 12, 100 copies per 25-μL reaction, with a median of 1, 645 copies per 25-μL reaction.
A total of 31 samples that had tested PCR positive for enterovirus were subsequently tested for viral infectivity using BGMK and Vero cells. No cytopathic effects appeared in BGMK and Vero cells at any dilution for all these samples following the infection period, indicating viruses eluted from samples, though present, were non-infectious under the present test conditions.
Six sites were sampled at two different time points to analyze both FIB counts and viral presence. There was very limited apparent correlation between FIB levels and viral presence in this study. Among the six sites tested for a total of 12 samples, only two freshwater sites and one brackish water site tested above acceptable limits for Enterococci, and no sites tested above acceptable limits for C. perfringens; however, 10 out of 12 sites tested positive for one or more enteric viruses(Table 3).
Table 3. Correlation between bacterial counts and viral presence determined in six sampling sites.
Sequencing and internal control
Table S3. Sequencing results of randomly selected amplicons detected during the seasonal study.