Comparison of Indicator Bacteria Concentrations Obtained by Automated and Manual Sampling of Urban Storm-Water Runoff

Abstract

A comparative study of indicator bacteria concentrations obtained by laboratory analysis of grab samples of storm water, which were collected manually or by automatic samplers, was carried out in two urban catchments. Samples were analyzed for four types of indicator bacteria, total coliforms, Escherichia coli (E. coli), enterococci, and Clostridium perfringens and further documented by measurements of total suspended solids (TSS) and turbidity. Analysis of complete data sets (N = 198) indicated no statistically significant differences in the geometric means of all the constituent samples collected automatically or manually, but there were some small differences between the results produced by the two sampling methods applied. Total coliform concentrations were positively biased in samples collected by automatic samplers, but for the three remaining indicator bacteria (E. coli, enterococci, and C. perfringens), the opposite was true. Risk of sample cross-contamination in automatic samplers was assessed in the laboratory by sampling consecutively synthetic storm water with high and low concentrations of E. coli and enterococci. The first low-concentration samples preceded by high-concentration samples were cross-contaminated and the measured concentrations were positively biased. This cross-contamination was explained by storm-water residue in the sampling line. Such a residue remained in place even after line purging by compressed air, and its mass depended on the sampling line length (tested up to 5 m), as verified by measurements in the laboratory. The study findings should be helpful for improving field protocols for indicator bacteria sampling.