ABSTRACT
Controlling the biological stability of water in distribution networks is a major concern of drinking water suppliers. Biodegradable dissolved organic carbon (BDOC) passing through the water treatment plant is one factor that contributes to biofilm formation. The release of growth-promoting compounds by certain materials coming into contact with drinking water has been de scribed as another contribution to microbial growth in drinking water distri bution systems (Schoenen and Wehse, 1988; van der Kooij, 1993). The ad vantages of the present study in investigating the early phase of biofilm development are: (1) the use of different kinds of pipe materials used in drink ing water technology, (2) the installation of devices at crucial points within the distribution system and (3) the use of different in situ investigations in combination with conventional techniques. The results demonstrated that plas tic materials were colonized in higher densities than the total bacterial cell count on steel and copper in bank-filtered water at all sampling points. The specific situation after the activated carbon filters at the waterworks may con tribute to the fast and increased colonization with bacteria. The biofilms on granular carbon particles are described as a monolayer facilitating an inten sive contact between microorganisms and substrates dissolved in water or fixed in carbon particle pores (Stringfellow et al., 1993; Davies and McFeters, 1988). This efficient nutrient supply may be responsible for an optimized metabolic physiology and an increased growth rate of bacteria. The high amount of metabolically active bacteria localized on released carbon fines, and also planktonic microorganisms in the effluent of the granular activated carbon filters, seem to be responsible for the rapid and dense colonization of the material coupons. In agreement with these results, Sibille et al. (1997) also measured a high amount of total fixed bacteria grown on PVC coupons in GAC water in the experiment (8.8 X 106 cells cm -2 ). Concerning the physiology of the attached bacteria, high dehydrogenase activities were observed in biofilms sampled after the filtration steps at the plant in this study. Organic compounds loaded on carbon fines, which attach to material surfaces, could support the nutrient supply of the biofilm bacteria, but also dissolved residues of dead mi croorganisms from carbon filtration could play an important role in nutrient supply. The disinfection with chlorine dioxide according to the German Drink ing Water Regulation (0.05-0.1 mg I-1 ) reduced the total cell count from test materials significantly. Nevertheless, physiologically active biofilms were de tected at the sampling point DIS, where the disinfectant was mixed with the conditioned water from the plant. Despite the effect of chlorine dioxide, a re tarded regrowth of biofilms was observed at the two house branch connection sampling points. Morin et al. (1996) and Servais et al. (1992) showed that bac teria associated with carbon fines could be carried through the disinfection barrier without injury. Therefore, these loaded particles that attach to pipelines
of the distribution system could be an important factor for biofilm regrowth. The adhesion of particles and microorganisms may also be influenced by the nature of the pipeline surfaces. Thus, chlorine dioxide induces corrosion of metallic pipe materials, increasing the release of corrosion products to the bulk water and roughening the inner surface of the pipes. Also, some bacterial species induce biological corrosion on various materials (Flemming and Geesey, 1991; Kaesche, 1990). All of these effects support the regrowth of the biofilms in the distribution system. For the present study the significance of these corrosion reactions are subordinated, because of the short exposure time of the tested pipe materials and low chloride dioxide concentrations. Similar to the biofilms developed in GAC water, the regrowth of the biofilms within the distribution system was again supported by plastic materials. The per centage of metabolically active bacteria compared with the total cell count in biofilms sampled within the distribution system was about 30-35% on aver age, which compares favorably with data from literature (Schaule et al., 1993). Therefore, the significant differences between the total cell count and meta bolic potential of biofilms from GAC to the distribution system may result from injuries to planktonic bacteria caused by disinfection. To compare con ventional plating methods with staining techniques, the results showed very clearly that the sensitivity of detection and the amount of information yielded from the DAPI/CTC assay for surpass conventional techniques. Kalmbach et al. (1997b) demonstrated that the inability of bacteria from oligotrophic wa ter systems to form colonies on commonly used media resulted from factors other than the nonviability of these bacteria. They postulated that the investi gated bacterial strains are present in distinct physiological states: (1) culturable, (2) metabolically active but incapable of undergoing the sustained cel lular division required for growth on artifical medium and (3) nonculturable and metabolically inactive.
