Another ramification of the presence of Klebsiella sp. is that
the organisms may have led to false positives for E. coli on the
chromogenic media used in this study. Large exports of both thermotolerant
coliform and bacteria that tested positive as E. coli were observed in all biofilter media that contained straw. A few studies have reported that some common non-E. coli thermotolerant coliform, including members of the Klebsiella genus as well as Enterobacter aerogenes, can produce the key enzymes, galactosidase and glucuronidase, that result in a positive E. coli detection on membrane filtration chromogenic media methods such as the Coliscan system.The non-specificity of these enzymatic detection systems wasobserved in this study as well, with Raoultella (Klebsiella) planticola and E. aerogenes accounting for roughly 70% of the presumptive E. coli colonies that were analyzed with genetic methods (Table 2).However, false positive rates for E. coli detection on chromogenic media are also known to be exacerbated by crowded plates, which were common in the WC/S trials because of large apparent export of Raoultella (Klebsiella)planticola and E. aerogenes from biofilters without accompanying
increases in serial dilution during monitoring. False positives induced by high levels of these bacteria at high concentrations on the Coliscan plates may partially explain the large E. coli levels observed in outflow from the WC/S biofilters. In the wood-based media trials, no E. coli were observed in outflow from SRA-WC mycofiltration media, however, low levels were observed in outflow from the VT-SRA-WC media. This monitoring indicated that there was not a native population of E. coli or false-positive-E. coli growing on the SRA-WC media; however, there may have been some export of resident E. coli or false-positive E. coli from the VTSRA-
WC media. This may partially explain the higher variability in E. coli removal rates seen in the VT-SRA-WC media relative to the SRA-WC media.