Although indicator organisms are often used as predictors of pathogen presence, pathogen concentrations can also be measured directly. Direct measurement is generally not done for routine monitoring efforts because of practical considerations; however, it was included in the Phase II design because historical pathogen-level data were lacking in the MMSD Corridor Study database.

The waterborne pathogens of interest during Phase II of the MMSD Corridor Study were bacterial, viral, and protozoal. Four common waterborne pathogens were chosen for direct measurement in Phase II: bacterial pathogens were E. coli O157:H7 and Salmonella and protozoal pathogens were Giardia and Cryptosporidium. Viral pathogen concentrations were not measured directly in Phase II but will be measured directly during Phase III of the MMSD Corridor Study.

Pathogens are generally present at substantially lower concentrations in water than indicator organisms are (Mara and Horan, 2003). As a result, detections are less frequent and concentrations are much lower than those associated with indicator organisms. Therefore, pathogen data are discussed in terms of detection frequencies (DF); if concentrations were reported for pathogens, the medians (where possible) and (or) ranges also are given. Because pathogen data were not present in the Phase I database, no comparisons of Phase I and Phase II data are possible.

Escherichia coli O157:H7

E. coli O157:H7 causes dysentery and hemolytic uremic syndrome in infected humans. The infective dose for this strain is believed to be fewer than 100 organisms (Percival and others, 2004). Reservoirs for this bacterium include humans and domestic animals. Because E. coli O157:H7 has not been shown to replicate in the environment, its presence is thought to be strictly related to fecal contamination. E. coli O157:H7 can survive in environmental waters for up to 21 days, but it is sensitive to chlorination treatment. Results for E. coli O157:H7 were reported as presence-absence and are discussed herein as number and frequency of detections.

E. coli O157:H7 was detected in 6 of the 237 samples collected during Phase II. All 6 detections were in samples from 5 stream sites (constituting 3 percent of samples). The site with two detections (DF of 17 percent) was the Menomonee River at Menomonee Falls site. The Underwood Creek, Milwaukee River at Mouth, Root River near Franklin, and Jewel Creek sites each had a single detection (DF of 8 percent). DFs indicated no appreciable relation to urban land use. Five detections occurred during low flow, and four of those occurred during summer.

Salmonella

Salmonella are a genus of common pathogenic bacteria that can cause gastroenteritis, enteric fever, and septicemia in infected humans. Sources of this group of bacteria are wild and domestic animals including cattle, swine, dogs, cats, birds, and humans. Salmonella have shown the ability to survive in environmental waters for prolonged periods; during warm months, they may be able to replicate in eutrophic waters. Although evidence suggests that Salmonella are less sensitive to disinfection techniques than coliforms, chlorination is effective in inactivating this group of bacteria. Previous studies have found Salmonella in 80 percent of activated sludge effluent from wastewater-treatment plants and in 58 percent of contaminated surface waters (Percival and others, 2004).

For Phase II of the MMSD Corridor Study, analysis for determination of Salmonella concentrations involved concentration enrichment and selective growth techniques, followed by serological testing and confirmation using polyvalent “O” antisera (U.S. Environmental Protection Agency, 2006a). Detectable concentrations of the bacterium were observed in 22 percent of samples. Given the large percentage of concentrations below the reporting levels (0.1 and 0.2 MPN/100 mL); detectable concentrations were considered too scarce to be analyzed in terms of overall medians. Data are discussed instead as the frequency of Salmonella detections in samples and the medians (where possible) or ranges in concentrations in samples with detectable concentrations.

return to top

Detection Frequencies

All 58 detections of Salmonella during Phase II were in stream samples (constituting 29 percent of the stream samples collected). No detections were observed in harbor samples, and this may be due to organism die-off or dilution when mixing with water from Lake Michigan. At least one detection of Salmonella was found at every stream site (fig. 38). Nine of the 15 stream sites had DFs greater than or equal to 33 percent; highest DFs (50 percent) were in samples from the Menomonee River at Menomonee Falls, Menomonee River at Wauwatosa, and Root River at Grange Avenue sites. DFs at remaining sites ranged from 8 to 17 percent. DFs indicated no appreciable relation to urban land use.

Figure 38. Detection frequencies and concentrations of Salmonella in samples, by site, in the Milwaukee Metropolitan Sewerage District planning area, Wis. Site abbreviations listed in table 1.

DFs indicated consistent responses in relation to flow and season. The DF for low-flow-event samples (23 percent) was much lower than that for high-flow-event samples (39 percent) (fig. 39A). The frequency of detection in spring samples (40 percent) was nearly twice as high as those throughout the rest of the year (23 to 25 percent) (fig. 39B). When flow was combined with seasonality, DFs for corresponding seasons were generally higher for high-flow events than for low-flow events (fig. 39C). The only exception to this general pattern was the frequency for autumn high-flow-event samples; this category, consisted of only three samples, none of which contained detectable concentrations of Salmonella. High-flow-event samples collected during spring (48 percent) and summer (40 percent) had the highest overall DFs, and, with the exception of the autumn high-flow-event sample, DFs for remaining high- and low-flow seasonal events ranged from 20 to 26 percent.

Figure 39. Detection frequencies and concentrations of Salmonella for stream sites, Milwaukee Metropolitan Sewerage District planning area, Wis. Stream samples are grouped by flow (A), season (B), and flow and season combined (C). Harbor samples are grouped by season only (D) (no harbor samples were collected in winter).

Concentrations

Salmonella concentrations greater than reporting levels were found only at stream sites. Concentrations ranged from 0.1 to 10 MPN/100 mL (fig. 38), and the median concentration was 0.2 MPN/100 mL. Samples containing the highest maximum concentration observed, 10 MPN/100 mL, were collected at eight sites: Willow Creek, Menomonee River at Menomonee Falls, Little Menomonee River, and Menomonee River at Wauwatosa, Milwaukee River at Mouth, Oak Creek, Root River at Grange Avenue, and Root River near Franklin.

There was little relation between the variability in Salmonella concentrations and flow or season. Differences between median concentrations for two flow-event conditions were very small, with medians of 0.6 MPN/100 mL for low-flow samples and 0.2 MPN/100 mL for high-flow samples (fig. 39A). The seasonal medians were highest in summer samples (10 MPN/100 mL) and the lowest in spring and winter samples (both 0.2 MPN/100 mL) (fig. 39B). When flow was combined with seasonality, highest maximum concentrations (10 MPN/100 mL) were found in samples collected during summer low- and high-flow events and winter high-flow events (fig. 39C).

return to top

Giardia

Giardia is a genus-level classification that refers to multiple types of protozoa. Only one species, Giardia duodenalis, infects humans. This species is a common waterborne protozoal pathogen that can cause severe diarrhea and malabsorption. Sources of Giardia include livestock, dogs, cats, beavers, guinea pigs, and humans. Giardia is released from host animals in the environmentally resistant cyst form. These cysts have been shown to maintain viability for up to 3 months in cold environments and to resist disinfection techniques that successfully remove bacteria and viruses. With the exception of response to ultraviolet radiation, Giardia cysts are generally more sensitive than Cryptosporidium oocysts. An infecting dose of 10−25 cysts has been observed in experimental infection studies (Percival and others, 2004).

For Phase II, analysis for Giardia concentrations involved filtration, flow cytometry, and fluorescent antibody microscopy (U.S. Environmental Protection Agency, 2006a). This method detects Giardia to the genus level only (U.S. Environmental Protection Agency, 2005a); therefore, reported results do not specifically refer to levels of Giardia duodenalis. Reporting levels for data in this data set varied greatly, ranging from 30 to 80 cysts/100 L. Concentrations above reporting levels were observed in 26 percent of the samples collected. Given the large percentage of concentrations below the reporting levels, detectable concentrations were considered too scarce to be analyzed in terms of overall medians. Data were reported instead as the frequency of Giardia detections in samples and the medians (where possible) or ranges in concentrations in samples with detectable concentrations.

Detection Frequencies

Giardia was detected in 27 percent of stream samples and in at least two samples (17 percent) from every site (fig. 40). The highest DF was in samples from Root River near Franklin (42 percent). Lowest DFs (17 percent) were in samples from Lincoln Creek, Milwaukee River at Milwaukee, Honey Creek, and Kinnickinnic River. DFs at remaining sites ranged from 25 to 33 percent. DFs indicated no appreciable relation to urban land use.

Figure 40. Detection frequencies and concentrations of Giardia in samples by site in the Milwaukee Metropolitan Sewerage District planning area, Wis. Site abbreviations listed in table 1.

DFs were related to flow and season. DFs at stream sites were slightly higher in samples collected during high-flow events (28 percent) than during low-flow events (26 percent) (fig. 41A). The highest DFs in stream samples were in winter (50 percent) and spring (32 percent); summer and autumn samples had identical DFs of 17 percent (fig. 41B). When flows were combined with seasonality, DFs at stream sites were not consistently lower during low-flow events when compared to high-flow events for corresponding seasons (fig. 41C); however, winter and summer samples had higher DFs than samples collected at other times of the year during the same flow condition.

Figure 41. Detection frequencies and concentrations of Giardia for stream and harbor sites, Milwaukee Metropolitan Sewerage District planning area, Wis. Stream samples are grouped by flow (A), season (B), and flow and season combined (C). Harbor samples are grouped by season only (D) (no harbor samples were collected in winter).

DFs at harbor sites (23 percent) were lower than those at stream sites (27 percent), and may be due to organism die-off or dilution when mixing with water from Lake Michigan. Inner-harbor sites had the highest DFs: the South Mid-Harbor Milwaukee Outer Harbor (OH-11) (50 percent), Middle Mid-Harbor Milwaukee Outer Harbor (OH-03) (50 percent), and North Mid-Harbor Milwaukee Outer Harbor (OH-04) (20 percent) sites (fig. 40). Giardia was detected in only one sample from among the outer-harbor sites (the Northern Outside Harbor Breakwall Lake site, OH-12, DF 11 percent). Harbor samples indicated a seasonal response. Spring samples had the highest frequency of detection (43 percent), followed by summer samples (12 percent), and autumn samples (8 percent) (fig. 41D). Samples were not collected from harbor sites in winter.

Concentrations

Giardia concentrations in stream samples ranged from 28.6 to 467 cysts/100 L, with a median of 60.6 cysts/100 L. Sites with the highest maximum concentrations were the Milwaukee River near Cedarburg (467 cysts/100 L), Milwaukee River at Milwaukee (375 cysts/100 L), Root River near Franklin (333 cysts/100 L), Willow Creek (329 cysts/100 L), and Root River at Grange Avenue (300 cysts/100 L) sites (fig. 40). Concentrations in stream samples indicated consistent responses in relation to flow and season. Stream samples had higher median concentrations during high-flow events (103 cysts/100 L) than during low-flow events (46.1 cysts/100 L) (fig. 41A). The highest median concentrations were in spring-season samples (121 cysts/100 L), followed by summer samples (66.7 cysts/100 L); autumn and winter medians were identical (33.3 cysts/100 L) (fig. 41B). When flows were combined with seasonality, the highest maximum concentration was in a sample collected during a spring high-flow event (467 cysts/100 L) (fig. 41C). Higher maximum concentrations were also observed in low-flow samples collected during autumn (333 cysts/100 L), summer (329 cysts/100 L) and spring (300 cysts/100 L), as well as summer high-flow samples (267 cysts/100 L).

Median Giardia concentrations at harbor sites (100 cysts/100 L) were higher than those at stream sites (60.6 cysts/100 L). Highest maximum concentrations were observed in samples from inner-harbor sites: Middle Mid-Harbor (OH-03) (200 cysts/100 L), South Mid-Harbor (OH-11) (133 cysts/100 L), and North Mid-Harbor (OH-04) (129 cysts/100 L) (fig. 40). The single detection in the outer harbor (at Northern Outside Harbor, OH-12) had a concentration of 33.3 cysts/100 L. Harbor samples indicated a seasonal response (fig. 41D). The maximum concentration in spring samples (200 cysts/100 L) was twice the maximum concentrations in summer and autumn samples (100 cysts/100 L in each).

return to top

Cryptosporidium

Cryptosporidium is a genus-level designation that refers to many protozoal species, some of which are pathogenic to humans. The primary pathogens, however, are generally considered to be C. parvum and C. hominis. Infection by Cryptosporidium can cause acute effects of diarrhea, abdominal pain, and vomiting; chronic effects include Reiter’s syndrome, a reactive arthritis. Reservoirs include humans and animals. Cryptosporidium species are shed in the environmentally resistant oocyst form, which has been shown to maintain viability for as long as 176 days in stream water. Oocysts are resistant to many forms of disinfection, including chlorination. The most effective means of inactivating Cryptosporidium oocysts is through the use of ozone or ultraviolet radiation. An infecting dose of 30 oocysts has been observed in experimental infection studies (Percival and others, 2004).

The analytical method for the determination of Cryptosporidium concentrations involved filtration, flow cytometry, and fluorescent antibody microscopy (U.S. Environmental Protection Agency, 2006a). This method detects Cryptosporidium to the genus level and does not have the specificity to distinguish between different pathogenic and nonpathogenic species; therefore, reported results do not specifically refer to concentrations of pathogenic Cryptosporidium (U.S. Environmental Protection Agency, 2005a). Reporting levels for data varied greatly, ranging from 28.6 to 233 oocysts/100 L; however, the majority of censored results (75 percent) were recorded as less than 33.3 oocysts/100 L. Concentrations above reporting levels were observed in 33 percent of the samples collected. Given the large percentage of results below the reporting levels, concentration data were considered too scarce to be analyzed in terms of overall medians. Data were reported instead as the frequency of Cryptosporidium detections in samples and the medians (where possible) or ranges in concentrations found in samples with detectable concentrations.

Detection Frequencies

Cryptosporidium was detected in 42 percent of stream samples and in at least two samples (17 percent) from every site (fig. 42). The highest DFs (both at 58 percent) were from Willow Creek and Menomonee River at Wauwatosa. The lowest DF (17 percent) was from Root River near Franklin. DFs at remaining sites ranged from 25 to 50 percent. DFs indicated no appreciable relation to urban land use.

Figure 42. Detection frequencies and concentrations of Cryptosporidium in samples, by site, in the Milwaukee Metropolitan Sewerage District planning area, Wis. Site abbreviations listed in table 1.

DFs varied by flow and season. DFs were higher in samples collected during low-flow events (46 percent) than during high-flow events (36 percent) (fig. 43A). The highest frequency of Cryptosporidium detections was in winter samples (70 percent) and the lowest in spring samples (18 percent) (fig. 43B). DFs ranged from 47 to 50 percent throughout the rest of the year. When flows were combined with seasonality, DFs were generally higher during low-flow events than during high-flow events for corresponding seasons (except spring) (fig. 43C). Within each flow category, the highest DFs were in winter samples and the lowest were in spring samples.

Figure 43. Detection frequencies and concentrations of Cryptosporidium for stream and harbor sites, Milwaukee Metropolitan Sewerage District planning area, Wis. Stream samples are grouped by flow (A), season (B), and flow and season combined (C). Harbor samples are grouped by season only (D) (no harbor samples were collected in winter).

The DF of Cryptosporidium at harbor sites (5 percent) was lower than at stream sites (42 percent), and may be attributed to organism die-off or dilution from mixing with water from Lake Michigan. Detections in harbor samples only occurred at the northern sites of the inner and outer harbor: North Mid-Harbor (OH-04) (20 percent, 2 samples total) and Northern Outside Harbor (OH-12) (11 percent, 1 sample total) (fig. 42). In relation to seasonality, the harbor samples had one detection per season, with DFs ranging from 4 to 8 percent (fig. 43D).

Concentrations

Cryptosporidium concentrations in stream samples ranged from 29.4 to 782 oocysts/100 L, with a median of 62.5 oocysts/100 L. The highest maximum concentration was at Willow Creek (782 oocysts/100 L) (fig. 42), and the lowest maximum concentrations (33.3 oocysts/100 L for both) were at Milwaukee River at Mouth and Root River near Franklin. Maximum concentrations at the remaining sites ranged from 66.7 to 286 oocysts/100 L.

Concentrations indicated consistent responses in relation to flow and season. Median concentrations during low-flow events (64.5 oocysts/100 L) were slightly higher than during high-flow events (53.3 oocysts/100 L) (fig. 43A). The highest median concentrations were measured in summer samples (66.7 oocysts/100 L) and winter samples (64.5 oocysts/100 L); spring and autumn samples had identical medians, 33.3 oocysts/100 L (fig. 43B). The maximum concentration for summer samples (782 oocysts/100 L) was much higher than the maximum concentrations in samples for the rest of the year (133 to 161 oocysts/100 L). When flows were combined with seasonality, the highest maximum concentration at stream sites was in a sample collected during a summer low-flow event (782 oocysts/100 L) (fig. 43C). High maximum concentrations were also found in samples collected during high-flow events in summer (233 oocysts/100 L), winter (161 oocysts/100 L) and spring (156 oocysts/100 L), as well as autumn low-flow-event samples (133 oocysts/100 L).

The two harbor sites where Cryptosporidium was detected had similar concentrations: North Mid-Harbor (OH-04) (32.3 and 33.3 oocysts/100 L) and Northern Outside Harbor (OH-12) (33.3 oocysts/100 L), which were generally lower than those at stream sites (median concentration of 62.5 oocysts/100 L) (fig. 42). Cryptosporidium concentrations were similar in spring, summer, and autumn (fig. 43D).

return to top
continue to Pesticides