A spatiotemporally static total coliform (TC) concentration threshold of five colony-forming units (CFU) per 100 mL is used in Ontario to determine whether well water is of acceptable quality for drinking. The current study sought to assess the role of TC and associated thresholds as microbial water quality parameters as the authors hypothesized that, since static TC thresholds are not evidence-based, they may not be appropriate for all well water consumers. A dataset containing the microbial water quality information of 795,023 samples (including TC and Escherichia coli (E. coli) counts) collected from 253,136 private wells in Ontario between 2010 and 2017 was used. To accurately assess the relationship between E. coli and non-E. coli TC, “non-E. coli coliform” (NEC) counts were calculated from microbial water quality data and replaced TC throughout analyses. This study analysed NEC and E. coli detection rates to determine differences between the two, and NEC:E. coli concentration ratios to assess links, if any, between NEC and E. coli contamination. Study findings suggest that spatiotemporally static NEC thresholds are not appropriate because seasonal, spatial, and well-specific susceptibility factors are associated with distinct contamination trends. For example, NEC detection rates exhibited bimodality, with summer (29.4 %) and autumn (30.2 %) detection rates being significantly higher (p < 0.05) than winter (21.9 %) and spring (19.9 %). E. coli detection rates also varied seasonally, but peaked in summer rather than autumn. As such, it is recommended that these factors be considered during the development of private well water guidelines and that static thresholds be avoided. Furthermore, the authors propose that, because NEC:E. coli concentration ratios change in the context of the aforementioned factors, they may have a role in inferring groundwater contamination mechanisms, with high ratios being associated with generalized aquifer contamination mechanisms and low ratios with localized contamination mechanisms.
Read full abstract