An investigation into parametric relationships between enterovirus and faecal indicator organisms in the coastal waters of England and Wales

Previous research has resulted in conflicting conclusions concerning possible statistical relationships between bacterial indicator and enterovirus concentrations in marine waters. Results have suggested either highly significant relationships or no significant correlation. To examine these relationships further, large compliance assessment data sets for identified bathing waters on the English and Welsh coast, between 1988 and 1992, were extended to include enterovirus and faecal streptococci counts in addition to total and faecal coliform data. The suitability of these data for parametric correlation analysis was assessed by examining the nature of the statistical distributions of raw data and residual values from regression analyses. The data were acceptable for parametric analysis when (i) zero enterovirus samples were excluded and (ii) enterovirus counts were logarithmically transformed. The analysis of over 690 data pairs produced positive Pearson correlation coefficients between coliforms, faecal coliforms, faecal streptococci and enterovirus concentrations. Though the correlations were significant (P < 0.001, i.e. > 99.9% confidence) they were low (r < 0.5). The amount of variance in enterovirus concentration explained by bacterial concentration was only 15–16%. Relationships were similar for each bacterial determinand. Correlations above successive quartile values of bacterial concentration revealed no threshold effect in these relationships. These results suggest poor statistical relationships between concentrations of bacterial indicators of sewage pollution and enterovirus in marine waters resulting in a low predictive capability.