Briefly, why was this Ohio beach research done, and what did you find?
We did the research to discover what environmental factors may be influencing concentrations of Escherichia coli (E.coli) bacteria in bathing waters at beaches and to determine whether the identified environmental factors could be used to predict E. coli concentrations in a timely manner. We found that we can enter current environmental and water-quality conditions—such as the wave height, cloudiness of the water, and rainfall—into an computer model for a particular beach and get a probability that the standard used to judge swimming safety will be exceeded given these condition. We developed models for three Lake Erie urban beaches, near Cleveland, and one inland-lake beach in Ohio.
How would use of the models differ from the current way of assessing swimming safety?
Current methods to assess recreational water quality rely on collecting a sample of water from the beach area, transporting it to a laboratory, and determining numbers of indicator organisms, such as E. coli. It takes as least 18 hours to grow E. coli in the laboratory. Therefore, by the time E. coli results are compiled, beach water-quality conditions may have already changed. By contrast, it takes less than an hour to provide an estimate of recreational water quality using the model.
Why is E. coli used to assess recreational water quality?
The E. coli bacterium, which is present in the feces of warm-blooded animals, is commonly associated with and is an indicator of several different disease-causing organisms, called pathogens. Viral pathogens are believed to be the major cause of swimming-associated diseases (Natural Resources Defense Council, 1998). There are more than 100 types of human pathogenic viruses that may be present in fecal-contaminated waters. Bacteria are true living cells and are larger than viruses. Waterborne bacterial pathogens in the United States include species in the genera Salmonella, Shigella, Vibrio, Campylobacter, and rare strains of E. coli. Protozoa are one-celled animals. The principal protozoan pathogens that are waterborne are Giardia lamblia and Cryptosporidium parvum.
Because of its association with pathogens and because it is comparatively easy to grow in the laboratory, E. coli is the basis for Ohio recreational water-quality standards. There are several varieties of pathogenic E. coli and hundreds of varieties of harmless E. coli. One variety of E. coli that is pathogenic has been associated with undercooked ground beef.
Is there a "pathogen problem" at Ohio beaches? Are people getting sick?
There is no true measure of the magnitude of disease associated with recreational water exposures (USEPA, 1999), in Ohio or elsewhere. That is because when people get sick after a weekend at the beach, the cause of illness is unknown. In addition, the illness is not reported unless there was an outbreak among a specific population, like a group of athletes participating in a triathlon.
Swimming in polluted water can make you sick (Natural Resources Defense Council, 1998). Epidemiological studies in the United States, however, have consistently found an association between gastrointestinal illness and exposure to contaminated recreational waters; the potential for other types of infections is not completely understood (USEPA, 1999).
What are the sources of pathogens at Ohio beaches?
Polluted runoff and stormwater; sewage spills and overflows (including combined sewer overflows, sanitary sewer overflows, and sewage from private sewage-treatment systems); and fecal pollution from birds, swimmers, or boats.
How did you collect the data needed for developing the models?
Lake-water samples were collected from two locations at each beach. After wading to knee-deep water, a sterile polypropylene bottle was opened about 18 in. below the water surface and filled.
Samples from the swash zone―the zone of the shoreline that is constantly washed by waves―were collected from one to three areas at each beach. In each swash-zone sampling area, two sets of three holes were dug with a sterile post-hole digger—one set 3 feet and a second set 6 feet inland from the outer edge of the swash zone. The hole was dug deep enough to allow standing water to accumulate. Once the water entered the hole, a sterilized well casing was inserted to prevent the hole from collapsing. We decided to sample the swash zone because a researcher in Indiana found that E. coli concentrations were elevated in swash-zone sands (Richard Whitman, U.S. Geological Survey, written commun.).
How well do the models work?
The models we developed are beach specific because the factors that affect recreational water quality are different at each beach. The models worked well at some beaches and not very well at others.
We tested the models using data collected during the recreational season of 2002. At Huntington and Villa Angela beaches, the models correctly predicted recreational water quality 70 and 94 percent of the time, respectively, and worked better than current methods. At Edgewater beach, the model correctly predicted recreational water quality only 48 percent of the time in 2002; the current method did better than the model. At the inland lake, Mosquito Lake, the model correctly predicted recreational water quality 76 percent of time; however, the model failed to predict poor water quality 4 out of 5 times the standard was exceeded. The current method also did a poor job in predicting when the standard was exceeded.
Although there is some uncertainty with using the models, the amount of uncertainty is quantified so that beach managers and the public can make informed decisions. We use the concept of a threshold. Probabilities equal to or above the threshold would indicate to the beach manager that E. coli concentrations are most likely above the bathing-water standard. Probabilities below the threshold would indicate that E. coli concentrations are most likely below the standard. In other words, our models can work like a weather forecast. If there was a 70 percent chance of rain, would you go to a picnic? A 20 percent chance of rain? Similarly, if there was a 70 percent chance that the standard will be exceeded, would you swim in the water? And again, because models provide an assessment of recreational water quality within an hour, they are better than the old methods that take 18 hours to complete.
Why was the USGS involved in this research?
Water quality and quantity are critical topics in which the USGS has developed expertise over many years. The relatively recent addition of biological science to our core programs gives us opportunities to apply science in a variety of new ways.
The USGS did not do this research alone. Agencies that cooperated with us were the Ohio Water Development Authority, Northeast Ohio Regional Sewer District, Cuyahoga County Board of Health, Cuyahoga County Sanitary Engineers, Cuyahoga River Community Planning Organization, and the Ohio Lake Erie Office. The Ohio State University (OSU), Great Lakes Forecasting System (GLFS) collaborated with us on the development of predictive models.
Is anyone else doing the same kind of research? Are other prediction tools being investigated?
Yes, The Cities of Milwaukee and Stamford, CT are doing similar research. The USGS Georgia District has initiated a bacteria alert program that provides the same type of output as our Lake Erie models—the probability that the standard will be exceeded. They have taken the results of their research one step further by posting a web site with model results (http://ga2.er.usgs.gov/bacteria/). Other researchers are investigating the use of rapid analytical methods for E. coli.
Will this research continue?
Because we believe this is a promising method, we will continue to seek funds for further research. The method was tested with data collected during 2002, so we need to also test it in subsequent years. Usually, models improve as more data are collected. Each year the model is tested, therefore, new data can be added and model variables can be recalculated to determine whether the predictive ability improves with an added year of data. If the model is able to predict recreational water quality as well as or better than use of antecedent E. coli for several years in a row, beach managers may consider using the models to aid or direct decisions on posting beach advisories.
Are some of the models ready to be used for beach advisories soon?
The models need to be refined and tested during future recreational seasons. If the models are able to perform well in predicating recreational water quality for 2 or 3 years in a row, beach managers and public health officials may chose to use the models in their beach notification programs. Setting up a beach notification system based on predictive models may require additional funding and resources.
USEPA states that a monitoring program is essential to any beach-management program and recommends that the current E. coli laboratory method be used for assessing ambient waters and for making decisions concerning the protection of human health (USEPA 2002, p. 4-17). However, USEPA also states that modeling tools may be used to supplement, not replace, monitoring. Modeling tools can provide conservative estimates when there is a lag time between sampling the water quality and obtaining results. If models are properly developed and applied, they can be used in making beach advisory or closing decisions (USEPA 2002, p. 4-22).
How can I find out more about your research? Our web site has information about this research (http://oh.water.usgs.gov/, click on Beach monitoring research). For general information on beach monitoring, go the USEPA Beach Watch web site (http://www.epa.gov/waterscience/beaches/).
Whom can I contact if I’m concerned about swimming at a particular Ohio beach?
You can find out who to contact about your beach by visiting the USEPA Beach Watch web site at (http://www.epa.gov/waterscience/beaches/).
Centers for Disease Control, 2003, Infectious disease information: Atlanta, GA, accessed February 2003 at http://www.cdc.gov/ncidod/diseases/index.htm.
Natural Resources Defense Council, 1998, Testing the Waters VIII: New York, 145 p.
U.S. Environmental Protection Agency, 2002, National beach guidance and required performance criteria for grants: Washington, D.C., EPA-823-B-02-004.
U.S. Environmental Protection Agency, 1999, EPA action plan for beaches and recreational waters: Washington, D.C., Office of Water, EPA-600-R-98-079.
For more information about USGS Beach-Monitoring Research, please contact Donna Francy, by email email@example.com or by phone 614.430.7769
|| Disclaimer || Accessibility
Last update: March 2003