E. coli has been demonstrated to be a more specific indicator for the presence of faecal contamination than the thermotolerant coliform group (Dufour, 1977). Improvements in both MPN and MF techniques have been carried out for the rapid and selective enumeration of E. coli. Barnes et al. (1989) designed a rapid seven hour membrane filter test for quantification of thermotolerant coliforms from drinking water samples and other freshwaters and salt waters, although it is not suitable for salt water due to the high proportion of false positives obtained. Fluorogenic and chromogenic tests using 4- methylumbelliferyl-β-D-glucuronide (MUG) have been applied in MPN and MF techniques, for the detection of β-glucuronidase that is specific to E. coli (Manafi and Kneifel, 1989; Balebona et al., 1990; Gauthier et al., 1991; Rice et al., 1991). A miniaturised MPN method with a 96-well microplate has been developed for E. coli (Hernandez et al., 1991; ISO, 1996b) (Table 8.3). Based on this principle a number of different media have been developed for the use in MF and MPN techniques (Frampton et al., 1988; McCarty et al., 1992). Commercially available media include Colisure (formerly Millipore, now IDEXX) (McFeters et al., 1995), Colilert (IDEXX) (Edberg et al., 1988; Palmer et al., 1993), m-ColiBlue (Hach), ColiComplete (BioControl), Chromocult (Merck) and MicroSure (Gelman). Similar media for the detection of E. coli in water have also been described (Sartory and Howard, 1992; Brenner et al., 1993; Walter et al., 1994). Molecular methods have also been designed to detect specifically E. coli from water samples, such as PCR-gene probes for the uid gene (Bej et al., 1991a,b; Tsai et al., 1993; McDaniels et al., 1996). In addition, other alternative techniques, i.e. enzyme capture (Kaspar et al., 1987) and radioisotopes (Reasoner and Geldreich, 1989) have been proposed.
Faecal streptococci and enterococci
Early attempts to quantify faecal streptococci relied on enrichment tube procedures and the MPN technique; Rothe Azide Dextrose broth followed by a confirmation in Ethyl Violet Azide (Litsky) broth being the procedure most widely accepted by researchers. A rapid system for enumeration of faecal streptococci or enterococci in water samples using a miniaturised fluorogenic assay based on a 96-well microplate MPN system has been described by several workers (Hernandez et al., 1991; Poucher et al., 1991; Budnicki et al., 1996) and the technique has recently been proposed as an ISO method (Table 8.7). In addition, Enterolert (IDEXX) is available for the MPN technique with up to 100 ml of sample, and has been shown to be reliable (Fricker and Fricker, 1996).
The enumeration of faecal streptococci by a MF procedure using a selective medium was first reported by Slanetz and Bartley (1957). Since then, several media have been proposed, including Thallous Acetate agar (Barnes, 1959), KF agar (Kenner et al., 1961), PSE agar (Isenberg et al., 1970), Kanamycin Aesculin Azide (KEA) agar (Mossel et al., 1973), mSD agar (Levin et al., 1975), and mE agar (APHA/AWWA/WPCF, 1989). The accepted standardised procedures for the MF method are given in Table 8.6. Other media formulations and incubation procedures for faecal streptococci have been proposed for specific situations (Lin, 1974), such as increasing the membrane incubation period from 48 hours to 72 hours to recover stressed faecal streptococci. Rutkowski and Sjogren (1987) developed a medium, designated M2, to distinguish between human and animal pollution sources.