The federal criterion for determining “safe” bacteria levels is based on studies conducted on large bodies of fresh water—Lake Erie at Presque Isle State Park near Erie, Pa. (above), and Keystone Lake near Tulsa, Okla. (right)—not on urban streams or creeks. Photos: Presque Isle State Park, Oklahoma Tourism and Recreation Department, and PBS&J
Houston's White Oak Bayou flood-control channel must meet the same stringent standard as that developed for beaches at recreational lakes.

Last March, the Texas Commission on Environmental Quality proposed that bacteria discharges in urban runoff to waters around Houston be reduced by 100%.

While committed to providing a healthy environment, stormwater managers expressed concern about the magnitude of the proposed reductions from wet-weather sources. Houston's Department of Public Works and Engineering urged the commission to review the proposed reductions, while Harris County and the Harris County Flood Control District suggested they might not be achievable given Houston's semi-tropical climate and frequent rain.

Like stormwater managers across the country, these public employees are struggling to reconcile a common dilemma: Urban creek restoration plans, known as total maximum daily loads (TMDLs), must achieve bacteria levels “safe” for swimming that were based on studies of large lakes—not creeks or streams. Because bacteria levels in urban runoff generally are elevated, discharge reductions can be large and challenging to achieve.

These infrastructure managers are working with state regulators to ensure that local instream standards and discharge reduction requirements are as realistic as possible. Their experience highlights the need to assess the technical issues underpinning federally mandated stream standards and to ensure that proposed discharge reductions address genuine water quality concerns.

Setting the Standards

In urban waterways, bacteria can come from many sources: storm-water runoff, illicit discharges, wildlife, leaking septic systems, sanitary sewer overflows, stream sediments, wastewater effluent, topsoil, and leaking sanitary sewer systems. Some contribute pollutants during dry weather and some during wet weather. Wet-weather sources of bacteria are particularly challenging to control.

The EPA's recommended freshwater criterion for bacteria, published in 1986, is based on studies conducted in freshwater lakes with designated swimming areas. From 1979 to 1982 the studies were performed during dry weather near wastewater treatment plant discharges. Because the study sites didn't directly receive stormwater runoff, the results don't reflect conditions in streams, rivers, or urban waterways. Yet these studies underpin the EPA's national water quality standard recommendation that state regulatory agencies use to establish safe levels in all fresh waters.

The studies assessed the concentration of three indicator organisms—fecal coliform, Enterococci, and Escherichia coli (E. coli)—to determine the organism most associated with illness. While the indicator organisms don't cause illness, they indicate the presence of fecal matter that could contain pathogens. The EPA measured concentrations of these organisms and obtained illness rates experienced by swimmers and nonswimmers on specific days. Armed with these data sets, EPA researchers used regression analysis to correlate indicator concentrations with illness rates.

Because the correlation between illness rates and E. coli levels was the highest, the EPA selected it as the indicator to be used to establish the water quality criterion for bacteria in waters designated for “contact recreation”—that is, swimming. After selecting an “acceptable” risk level of eight illnesses per 1000 swimmers, the agency established a freshwater criterion requiring that the geometric mean density of E. coli not exceed 126 colonies per 100 milliliters of fresh water.