Water Quality Management - River Dynamics and Integrated River Management - page 599

based on concentrations of total coliform and fecal coliform bacteria, because there are many environmental
sources (e.g., soil, water, birds, sand, plants) for total coliform and fecal coliform. Therefore, the
concentrations of these two groups of fecal bacteria in environmental waters often do not necessarily
represent human fecal contamination. However, enterococci and E. coli also have environmental sources
(soil, plant, and water), so the same argument the USEPA used to reject total coliform and fecal coliform
could be used to refute the usefulness of enterococci and E. coli indicators (MWRDGC, 2006). These
extraneous sources of indicator organisms best explain why even beaches in relatively undeveloped areas
routinely exceed USEPA-recommended recreational standards, such as the beaches in Door County,
Wisconsin monitored by Bower et al. (2005). In general, the position of the regulated community is that
the 1986 USEPA criteria replaced an imperfect indicator with two other imperfect indicators, and that
more studies are required (MWRDGC, 2006).
A main shortcoming of the 1986 USEPA criteria is that in the epidemiological studies, the strength of
the association between swimming and illness at the freshwater sites was weak—only 2 of 9 trials
showed statistically significant difference between swimmers and controls (non-swimmers) [MWRDGC,
2006]. This may be reflective of the insufficient power of the studies (too few subjects), or high levels of
illness in the non-swimming control individuals in the studies. That is, in none of the epidemiological
studies was a control group used that did not visit the beach, which can lead to bias in that the beach may
have caused the illness not just direct contact with the water. Further, the focus of the epidemiological
studies used by the USEPA in the development of criteria (and basically all such studies available) has
been on gastrointestinal illness without considering water-contact diseases.
Another main shortcoming of the 1986 USEPA criteria is that they are not truly based on the human
health risk posed by contact with the water. In the epidemiological studies, no attempt was made to
evaluate the actual level of exposure (i.e. length of time in the water, extent of contact time with water,
amount of water ingested, etc.) of each swimmer at the beaches studied, nor was any attempt made to
evaluate the relationship between levels of indicators and actual pathogenic organisms at the study sites
(MWRDGC, 2006).
More recently it has been found that the foundation of the indicator organism approach—namely, the
implicit assumption that an approximate trend exists for human pathogens to be at higher levels whenever
indicators are at higher levels—may be faulty. Recent studies indicate that there is poor correlation
between indicator bacteria levels and levels of human pathogenic bacteria, viruses, and protozoa (Noble
et al., 2006; Noble and Fuhrman, 2001; Hardwood et al., 2005; Jiang et al., 2001; Hörman et al., 2004).
For example, Bower et al. (2005) found human specific Bacteroides spp. (indicative of human fecal
contamination) when culturable E. coli levels were as low as 30 to 105 CFU/100 mL in nearshore Lake
Michigan and 110 to 170 CFU/100 mL at a Lake Michigan beach following a combined sewer overflow
in Milwaukee, U.S. However, Bower et al. (2005) also found that for surface runoff with high (> 10,000
CFU/100mL) E. coli levels Bacteroides spp. was present, but the human-specific genetic marker was
negative. Results such as the former have led to concern regarding whether water body safety standards
based on indicator bacteria adequately protect human health. Whereas results such as the latter have
shown that proposals to disinfect urban stormwater runoff to meet fecal coliform or E. coli standards may
be unnecessary. Such a proposal to disinfect urban stormwater runoff was included in a draft of the
Southeastern Wisconsin Water Quality Management Plan Update (SEWRPC, 2007), but it was replaced
by a program ķ to sample for human specific Bacteroides in stormwater runoff and to remediate leaky
storm sewers and ĸ complete a microbiological human health risk assessment (Section 9.4.4).
Human sources of fecal pollution constitute a serious health risk because of the high likelihood of the
presence of human pathogens. Thus, if an indicator could be determined that is directly linked to human
sources, it could be a powerful tool for determining which water bodies need remedial action with