Jackson Park home.
Page on Swim bans, efforts at better testing and remediation, the new standard for swim bans and public notification.

Characterization of E. coli Levels at 63rd Street Beach

(The 'Whitman Report'), December 3, 2001

Summary and key sections, with annotations by Gary Ossewaarde


Characterization of E. coli Levels at 63rd Street Beach

Prepared for
City of Chicago
R. L. Whitman, T. G. Horvath, M.L. Goodrich and N.B. Nevers,
U.S. Geological Survey
Lake Michigan Ecological Research Station
1100 N. Mineral Springs Rd.
Porter, Indiana 46304

M. J. Wolcott
U.S. Geological Survey
National Wildlife Disease Center

S. K. Haack
U.S. Geological Survey
Water Resource Division, Michigan District

December 3, 2001



At the request of Chicago Mayor Richard M. Daley, the Chicago Department of Environment, Chicago Park District and the United States Geological Survey (USGS) developed a cooperative relationship in early 2000 to provide information on the nearshore waters of the Chicago area. In particular, the City wanted a thorough understanding of excessive E. coli occurrences at 63rd Street Beach and to apply this knowledge to all of Chicago’s beaches. In 1999, 63rd Street Beach had high E. coli levels, as defined by federal water quality and state guidelines for swimmable waters,[*] significantly more often than other Chicago beaches, leading to more restrictions on swimming. In the interest of public health and recreational enjoyment, Chicago has one of the intensive monitoring programs in the Great Lakes region. The City voluntarily complies with and often exceeds federal and state guidelines in order to ensure public safety when swimming restrictions occur.

[* Page 7 Resources, states under Park District that “When E. coli levels exceed the U.S. Environmental Protection Agency’s (USEPA) safe standard for swimmability, the CPD is responsible for closing swimming areas and informing the public when swimming is being restricted.” Under USEPA, it says “The USEPA is responsible for setting the safe standards for public health on beaches. Their scientists have conducted extensive research on E. coli as an indicator of pathogens harmful to human health. The City involved the USEPA in reviewing the City’s water sampling, testing and monitoring procedures.”

My (Gary Ossewaarde: conclusion: the standard being used by the city in 2001 (235) was what EPA then had in place and based on EPA “extensive research.” What changed? On page 8 it also says IDPH “provided valuable information on testing protocol (both old and new) and conducted tests on water samples collected throughout the 2000 swimming season.” Jardine Treatment Plant did the actual city/CPD tests.]

[Return to text, p. 6.]
Chicago’s intent with this study was to gain a thorough understanding of the nature and source of E. coli problems and begin to address remediation. Specifically, Chicago wanted to identify sources that may be contributing to elevated E. Coli levels and quantify to what extent these sources contribute to the E. coli levels at 63rd Street Beach. Further, Chicago desired information on factors that related to high E. coli levels to evaluate the utility of developing real-time prediction model that could augment and even replace traditional monitoring approaches. USGS desired the opportunity to gather intensive information on the spatial-temporal distributions and population characteristics of E. coli in order to understand how environmental conditions, sources, and bacteria concentrations interact. This in turn would help USGS provide information for better management of public swimming areas in the Great Lakes area.

A team of experts and support staff was assembled…. [institutions included] Lake Michigan Ecological Research Station; Great Lakes Science Center; National Wildlife Health Center; Michigan Water Resources Division, USGS; and Indiana University.

This report goes far in achieving the goals set above. It states that the origination of E. coli is very complex and sources are varied. It discusses some of the specific factors contributing to high E.coli levels and some hypothetical options tat could alleviate or manage high E. coli concentrations.

The Chicago 63rd Street Beach project has underscored the knowledge that even a small system, such as 63rd Street Beach, is extremely complex and varied. The study employed many of the available modern scientific tools (intensive descriptive statistics, modeling, advance biochemistry, genetic fingerprinting, biotyping, antibiotic resistance profiling, and still it is difficult to determine all the factors affecting E. coli on the beach. No single factor, or even a set of factors, emerged as being universally reliable to predict specific concentrations of E. coli. The study says a lot about how the system does work; provides more data than ever about the system; reveals patterns in the natural system that may contribute to beach management strategies, such as the timing of tests and beach maintenance; and has clearly laid the foundation for further studies both in the interest of Chicano and the Great Lakes.

[Page 9, Summary]



To characterize the distribution and possible sources of E coli at 63rd Street Beach, Chicago, an intensive study was undertaken between April and September 2000. Swimmability has been affected by high concentrations of E. coli in the past several years and in particular during the summer of 1999. Beach closures are enforced to protect the public from possible harmful illnesses associated with contamination. Most strains of E. coli are harmless, but it is typically associated with more harmful bacteria that can cause illness. The City of Chicago wanted to eliminate E. coli contamination at the beach in order to increase swimming safety and reduce beach closures. In order to accomplish this, sources of E. coli and the movement of E. coli within the system had to be determined.

Daily Sampling
Over the course of six months, water samples and sand samples were collected. In April, water samples were collected at two depths (45 cm and 90 cm) along five transects three days each week, and onshore and submerged sand samples were collected in these transects. Additional water samples were collected off the north revetment, at the end of Casino pier, near the mouth of the Jackson Harbor, and from the Jackson Lagoon outflow. One additional sand sample was sample was collected where the density of seagulls on the beach was observed to be the highest. Between May and September, an additional set of water samples was collected in the afternoon at the same locations along the five transects. Field observations were also noted in the morning, including the number of gulls on the beach, wind speed and direction, air and water temperature, and wave height at 45 cm depth.

Replication and Hourly Sampling
On ten randomly selected days, replicate water samples were collected. Two water samples were collected at each sampling site, and ten samples were collected at the 90 cm site in one transect. During ten other randomly selected days, sample were collected at the usual ten sites hourly from 7:00 a.m. to 3:00 p.m.

In order to test a hypothesis about E. coli survival during normal sunlight exposure, an experiment was conducted on-site using clear and dark bags containing lake water. The experiment was conducted on September 18 between 8:00 a.m. and 3:00 p.m.

Groundwater Testing
Seepage meters and piezometers were deployed on two separate days to determine the direction of water flow between the beach and the lake. Seepage meters were placed either in the lake bed or in the beach swash zone. Water movement was recorded, and samples were collected for E. coli analysis.

Source Testing
Additional tests conducted included DNA analysis of gull droppings to determine potential sources of E. coli at the beach. Water, sand, and fecal samples were analyzed for E. coli and Salmonella ssp. using rep-PCR and pulsed gel electrophoresis. Isolates were also analyzed for antibiotic resistance of E. coli and Enterococci.
An analysis of harbor water and sediment was conducted once to determine potential E. coli sources. Samples were collected in Jackson Harbor and analyzed for E. coli concentrations.
Finally, water samples from each transect and the lagoon were tested for wastewater compound in order to determine potential sources of E. coli. The water was analyzed for chemicals that would indicate human influence.

Weather data were collected throughout the study at a weather station located on Casino Pier. Data collected included wind speed and direction, barometric pressure, temperature, rainfall, and solar radiation.
Ambient water conditions were tested simultaneously throughout the sampling period using a multiprobe water monitoring instrument. Temperature, pH, conductivity, dissolved oxygen, turbidity, chlorophyll a, nitrate, and ammonium were measured every fifteen minutes at a remote platform located in 1.25-m-deep water. Using these data and E. coli results from water and sediment sampling, models were developed to predict elevated E. coli concentrations at the beach.


Daily Sampling
E. coli concentrations in water samples at both depths and times collected were correlated with each other, and similarly, E. coli in sand samples at foreshore and submerged sites were correlated. Comparing the two water depths, E. coli concentrations were considerably lower in the deeper water (90 cm) than in the shallow water (45 cm), and counts in the offshore water (off the pier) were lower than both shallow (45 cm) and deep (90 cm) water. Overall, E. coli concentrations were higher in morning water samples than in afternoon samples. Overall, E. coli concentrations were highest in sand samples collected near the highest density of seagulls on the beach. Time of day and location of collection are clearly important considerations for beach monitoring with the amount of variation found in this study.

Replication and hourly Sampling
Results of replicate sampling indicate that a single sample is not sufficient for accurately estimating E. coli concentration in the water—the techniques used by most E. coli monitoring programs.
Hourly sampling results indicate a dramatic decrease in E. coli concentration over the course of the day. E. coli concentrations exceeding the safe limit in the morning typically dropped off to concentrations below the safe level in the afternoon. On days when samples were collected twice, afternoon samples were significantly lower than morning samples. The samples collected in 90 cm of water showed a smoother decreases over the course of the day than water collected from 45 cm depth. Samples collected on the ten instances of hourly sampling extended this result. E. coli concentrations decreased exponentially between 8:00 and 15:00.

Light readings clearly supported the hourly sampling results. Over the course of the day on September 18 and 25, both visible and UV increased between 7:30 and 13:00 and then appeared to fall off slightly. Submerged probes indicated that water severely impeded UV penetration, more so than visible light. Results of the light/dark bag experiments supported the hourly sampling E. coli results. E. coli concentrations decreased throughout the day in the bags exposed to sunlight and in ambient lake water while concentrations in bags shielded from light decreased only slightly through the day.

Groundwater Testing
Groundwater studies indicated that the general movement of water was downward into the sand except for the swash zone, where the gradient was directed from the sand to the lake. Seepage flux was always limited, but the E. coli concentrations in seepage water were highly variable.

Source Testing
Bird density and location was also compared with E. coli concentration to examine any correlations. In the morning, gull numbers increased in May to peak in July and then began to decrease. In the afternoon, gull numbers increased consistently from May to September. During both time periods, gulls typically occupied the north end of the beach. Lagged bird counts were correlated with sand and water. Number of bathers was not correlated with E. coli concentrations in the water.
Jackson Harbor sediments and water apparently were not important sources of E. coli to the beach because concentrations were relatively low. Lagoon water concentrations were also low. Seagulls are a source of E. coli, but other sources are possible. Fingerprinting of seagull DNA isolates indicated that E. coli and Enterococci at the beach were partly derived from resident seagull population DNA analysis of Salmonella spp. indicated a relatively close math between gull droppings, water and sand samples, but some Salmonella spp. could have been transferred from other birds. E. coli and Salmonella were both highly susceptible to antibiotics, indicating a non-human source. These results were supported by the chemical analysis of water samples. Although numerous anthropogenic biochemical were present, they are likely derived from storm and wastewater rather than sewage.


Efforts to model the occurrence and prevalence of E. coli met with some success. There were correlations between elevated E. coli levels and storm events and the associated high winds and waves. There was no single factor that could be used to predict accurately the concentrations of E. coli. The best predictors overall were rainfall, wave height, wind speed (northern component only), air temperature and solar radiation, lake stage (level), water turbidity, an chlorophyll a concentration in the lake water.


The goal of this study was to determine the potential sources and distribution of E. coli at the 63rd Street Beach. Because of the unique structure of the beach, it is likely that E. coli may be moving south into the area of 63rd Street beach, where it becomes trapped due to shallow depths and the presence of a large pier. With such a scenario, E. coli levels could originate from any number of sources along the Chicago lakefront. Although all sources were not identified in the course of this study, it was determined that seagulls and sand E. coli are among the largest contributors.

In order to protect beach visitors from high E. coli levels, sources ultimately need to be determined and eliminated. Because the problem still persists, personnel must continue to monitor the beaches for excessive concentrations of E. coli. Statistical analysis shows that ten replicate samples are are needed to get a reliable indication of E. coli concentration. Nevertheless, the Chicago Park District’s monitoring exceeds national standards for testing. Predictive models that were developed over the course of this study may alleviate shortfalls in sampling precision and timely reporting. Using ambient conditions, a model was developed that can predict excessive E. coli contamination most of the time. We suggests a more comprehensive validation and calibration of this model in 2001.

The complexity of the 63rd Street Beach system and the interacting factors associated with a beach in a metropolitan area make source determination difficult. The results of this study illuminate some of the factors and eliminate others. with more information about other beaches and influences along the Lake Michigan shoreline, E. coli levels may eventually be minimized.

[Summation of scenario: beach and pier geography trap E coli from the north, which becomes trapped and entrained in the water and sand and birds. Gulls are a major factor, as are weather conditions and events, but apparently not humans except for a chemical pollution component from runoff.

Testing, to be reliable, ought to be more intense, but predictive models can predict the contamination and spiking most of the time.]


[Pages 13-83:
Methods (Study Area, Collection, Field Observations, Sampling Schedule, Dark/Light Bag, Additional Harbor Testing, Ambient Water Conditions, Test of E. coli Concentration, Groundwater Studies, DNA fingerprinting, Antibiotic Resistance Testing of E. coli and Salmonella, Chemical Tests for Wastewater Compounds, Statistical Data Distribution.

Embedded conclusions:

Foreshore sands and shallower(45 cm) water had significantly higher concentration of E. coli, with the sand having significantly more.

Implications for monitoring are great—time of day has great effect on abundance (much higher in the morning; that’s when sampling should be done, but sampling should not be done just once day). But how deep you sample also matters—but the shallower varies more.

Concentrations usually drop exponentially through the day (less robust at shallower depths) as light kills E. coli, except on a few days, more likely to be when the level was already low.

Water coming from the shore sand into the swash zone (not a large component) varied a lot in concentration, with one major spike August 22 (orders of magnitude). A stormy period also had a high concentration in incoming wastewater. However, there was very weak correlation between level in the lagoon and in the submerged sands.

Raking gull dropping seems to increase the bacterial concentration in sand and water. Wider beach sections have more gull droppings and bacteria levels.
Number of gulls is positively correlated with E. coli concentrations in foreshore sands and inshore water. (Influence of the gull population lags a day.)

Changing sand to 18 inches seems to have reduced E. coli for a while (not robust for other factors), but within two weeks was contaminated as before.

Wave height is one of the most important and revealing factors influencing E. coli. High waves result in hydrologic energy in the swash zone, which mobilizes and suspends E. coli-laden exposed or submerged sediments, and if high enough engages the beach sand and gull droppings.

P. 58: Morphology and Remediation
Beach and adjacent morphology may be an important aspect of water quality at 63rd Street Beach. Morphology directly affects circulation patterns, waves, sediment deposition, resuspension, and entrainment of contaminants and helps control the export or dilution of pollutants such as E. coli. The bounding of 63r Street Beach by the Jackson Harbor breakwater and the doglegged Casino Pier forms an embayment. To develop a confident understanding of the hydrodynamics of the 63rd Street Beach ‘embayment’, circulation tracing and hydrodynamic models would have to be employed. Nevertheless, there are some approximate intuitive assumptions that can be made by the morphological settling at the beach. With a northerly longshore current created by high winds, suspended contaminants might be captures by Casino Pier. Because of the geometry of he embayment, wind-generated internal circulation during calmer periods might be favored, thus retarding exportation, dilution, and externally and internally introduced E. coli. Sand and fine sediments with associated E. coli become trapped in the embayment. Accrual of sediments accounts for the beach’s shallowness, the increased silty nature of the sand and the more abundant natural and anthropogenic debris along he beach relative to other observed nearby beaches. The north and south breakwaters also act as protection against wind by reducing fetch. The increased calmness may translate itself into reduced exportation and dilution f internal or external loadings of bacteria and anthropogenic chemicals that were found in the study. On the other hand, the walls increase the total energy in the area during certain flow conditions because wave energy is reflected rather than absorbed. This energy may act to keep material suspended long enough for some of it to be exported from the beach area. The shallowness of the beach compounds the problem by further decreasing circulation and by allowing less volume of water for dilution of bacteria. Bacteria tend to be associated with detritus and fine sediments (i.e. silts and clays). In deeper water this bacteria-laden material eventually settles to the bottom, and the bacteria eventually dies. This may be why offshore water, and even 90-cm water or harbor water, was lower in E coli content than 45-cm water.

….increasing circulation could be desirable. These hydrodynamic questions lie beyond the scope of the current study but need to be answered before significant funds and effort are made towards remediation. Some of the remediation options that might be considered are: 2) dredging just beyond the swimming area to increase circulation and increase volume, 2) allowing water to move more freely through Casino Pier or 3) creating a sand shunt under the Pier. …

Number of humans in the water appeared to have no effect, and the pathogens tested did not have the resistance those from humans have.

Yet, (p. 75) there is significant contributor other than gulls to the bacteria—at times to the great majority.

And other contributors to and types of pollution in the water:

Evidence for other sources
Water samples were analyzed for a variety of constituents associated with human wastewater…These constituents include a variety f chemicals used (detergents, fumigants, solvents, polycyclic aromatic hydrocarbons, an insect repellant, fire-retardants) consumed (caffeine, cotinine—a byproduct of nicotine); or produced (cholesterol, coprostanol) by humans. These constituents commonly occur in water influenced by human activities through the addition of storm water or sanitary waste. Results indicate that beach water contained some of these target chemicals on September 11, 2000. This sampling date followed a significant rainfall event. The results suggest that storm water runoff affects the beach water under such conditions. The chemicals detected are not especially suggestive of human sanitary wasted since the detergents and human metabolites such as cholesterol and coprostanol were not detected. Rather, the combination of chemicals detected could conceivably arise from water washing off parking lots or picnic areas.

…Other sources may include
· Storm-water runoff
· Bacterial associated with aquatic plants or detritus brought to the beach by long-shore drift
· Bacteria associated with septic or sewage waste brought to the site by long-shore currents that may originate from gray water waste from boats or domestic effluents
· Groundwater or through-flow inputs (e.g., seepage from the lagoon, or harbors or washing-in of bacteria on the “hill” separating the lagoon and beach which would then be transported to the shallow water table.

The hydrologic connection between the lagoon and/or harbor and specific sites on the beach through human infrastructure and geologic heterogeneities also may present potential contamination channels. In addition, there may be current that affect the movement of water along the beach and a may provided unanticipated connections between sites. Since the chemicals found are generally persistent, they may have been in the system for quite some time with origins that are relatively distant to Chicago.
…The presence of chemicals indicative of human-influenced storm or wastewater in beach waters is significant. This is the first report of such detections in ambient lake waters and may indeed be common at many urban beaches. However, this influence will have to be taken into consideration as mitigation strategies to improve recreational water quality at the 63rd Street Beach are undertaken….

Some correlation of far-out water bacteria (with some resistance) to that of the lagoons was noted but inconclusive. Multiple antibiotic resistance (indicative of human agency bacteria and Salmonella) was not found.
Gulls were found to be a potent source of Salmonella at the beach

Two pages of technical references