Using Bacterial Source Tracking and Other Innovative Techniques to Identify Sources of Fecal Contamination in Stormwater

Stormwater runoff can transport a variety of pollutants including heavy metals, excess nutrients, sediments, trash, etc. Several studies dating back to the 1970s have shown that a number of human pathogens are often present in stormwater and can lead to serious health risks. The resulting health and economic impacts can be significant since the presence of elevated concentrations of microorganisms can lead to closures of shellfish harvesting areas and recreational beaches. Exceedence of microbiological standards is a common Total Maximum Daily Load (TMDL) impairment not only in Florida, but also throughout the U.S. Unfortunately, the process to directly enumerate disease-causing bacteria and viruses is time-consuming and expensive, making it virtually impossible to test for all possible pathogens in a water sample. As a result, certain types of bacteria (typically total and fecal coliform bacteria and/or enterococci) are used by regulatory agencies as indicators of microbiological water quality. Unfortunately, the fecal coliform standard has often been disputed with respect to its predictive capabilities since several recent studies have shown that this group of bacteria is capable of surviving in tropical climates outside its human host. Other studies have shown that it may not always be present in conjunction with other microbial pathogens. A number of relatively new techniques have been developed to assess microbiological water quality including antibiotic resistance pattern analysis, gene probes, polymerase chain reaction (PCR), and biosensors. These various techniques will be discussed, including their strengths and limitations and applicability for use in stormwater evaluations and development of best management practices. INTRODUCTION Certain types of bacteria are used by regulatory agencies (U.S. Environmental Protection Agency, Florida Department of Health, Florida Department of Environmental Protection) to assess microbiological water quality. The indicator bacteria are intended to act as warning signals that the water may be contaminated with feces from animals and/or humans. The presence of fecal material in water increases the likelihood that humans who drink, swim in, or Kurz, Harwood, Rose, and Lim Seventh Biennial Stormwater Research & Watershed Management Conference May 22-23, 2002 61 consume uncooked shellfish from those waters will contract a waterborne disease. Regulatory agencies routinely use the fecal coliform group of bacteria as a fecal indicator. The current standards for fecal coliform bacteria and several other indicators of fecal pollution are presented in Table 1 below. Enterococci has recently been adopted by the U.S. EPA as the preferred indicator organism for marine waters. Table 1. Guidelines for indicators of fecal pollution in surface waters. Parameter Guideline Fecal Coliforms EPA and State of Florida recommended guideline for a single sample not to exceed 800 cfu/100ml or a monthly geometric mean of 200 cfu/100ml. E. coli EPA recommended guideline for a geometric mean of 126 cfu/100ml. Enterococci EPA recommended guideline for a single sample of 104 cfu/100ml or a geometric mean of 33-35 cfu/100ml for marine and freshwater, respectively. Clostridium perfringens Guidelines based on University of Hawaii study (Fujioka) of 50 cfu/100ml for fresh/brackish water and 5 cfu/100ml for marine waters. Coliphage Guideline based on University of South Florida studies (Rose) of 100 pfu/100ml. The goal of bacteriological water quality testing is to predict the risk of disease based on measured levels of bacteria and/or bacterial products. This goal has been elusive, in part, because of the limitations of existing, approved methods. It is difficult, time-consuming and expensive to directly enumerate all of the potential disease-causing bacteria and viruses that may be present in a water sample. As a result, resource management agencies typically measure the numbers of indicator bacteria, whose presence more or less reflects the probability pathogens are present in a given waterbody. However, the fecal coliform indicator is a poor predictor of viral pathogens, and may well be present in waters where there are few or no pathogens of any kind (viral, bacterial or protozoan). Although the enterococci may be better predictors of viral pathogens in some areas of the country, they may also be present when pathogens are absent. One of the major reasons that fecal coliforms and enterococci are inadequate indicators is that they are present in the gastrointestinal tract of all warm-blooded animals. Some animal feces, i.e. those of humans, cattle, and swine, have a higher probability of containing human pathogens than the feces of most other species, therefore these animals are included in the “high risk” group. Very low levels of fecal indicator bacteria from a high risk animal group would indicate a greater potential health hazard than higher levels of indicator bacteria from a low risk animal group. Currently, there is no routine testing method that can be used to determine the origin of fecal indicator bacteria, however such a method would allow much more accurate risk assessment than that which can be achieved with standard testing methods. This would also allow regulatory agencies to more effectively identify and eliminate the source of bacterial contamination, which could lead to the reopening closed shellfish beds, and fewer health advisories postings at recreational beaches. Kurz, Harwood, Rose, and Lim Seventh Biennial Stormwater Research & Watershed Management Conference May 22-23, 2002 62 Accurate detection and identification of fecal contamination in surface waters (including stormwater) is a critical component of the federal Total Maximum Daily Load (TMDL) initiative. At the state level, over 15% of the reported waterbody impairments in Florida (303[d] list) were due to exceedences of fecal coliform bacteria concentrations. However, it is not known, definitively, whether these exceedences represented actual contamination of organisms capable of causing diseases in humans. This presentation will include a discussion of several newer methodologies and technologies to identify sources of fecal contamination in surface waters (and stormwater). A methodology for the detection of fecal contamination problems and a diagnostic process to identify potential sources will also be discussed along with potential solutions (BMPs) to reduce concentrations of pathogens. METHODS Fecal Contamination Assessment A case study using the Hillsborough River watershed is presented as an example of a fecal contamination assessment followed by a bacterial source tracking study (Kurz and Harwood, 2001). Historical fecal coliform concentrations in the upper Hillsborough River have exceeded state standards on numerous occasions. This frequent exceedence of water quality standards was identified in the SWFWMD’s Hillsborough River Comprehensive Watershed Management Plan as an issue requiring further evaluation. In order to identify the most fecally contaminated areas of the watershed, data were collected for the project area through the Hillsborough River Watershed Management project performed for Hillsborough County’s Stormwater Management Section. Sources of data were from STORET, the Hillsborough County Environmental Protection Commission (HCEPC), and Ayres Associates (for water quality sampling performed during the year 2000). Within the Hillsborough River study area, analyses of recent data indicated that the Class III standard of 200 cfu/100ml is exceeded at HCEPC stations 108 and 143 within the Blackwater Creek region as well as stations 118 and 148 representing the adjacent Lake Thonotosassa/Pemberton/Baker Creek drainage area (Figure 1). A more detailed evaluation of monthly trends between 1988 and 2000 for stations in the Blackwater Creek area are shown in Figure 2. Trends in fecal coliform concentrations have generally declined since the late 1980s. One reason for this trend may be due to the closing of several dairies in the watershed (Richard Boler, HCEPC, pers. comm.). However, concentrations at both stations frequently exceed both the 200 and 800 cfu/100 ml Class III standard which has resulted in the closure of the upper Hillsborough River to swimming/contact activities. A number of recent studies have investigated methods for predicting trends in fecal indicators including a study conducted by Lipp et al. (in press) and McLaughlin et al. (in prep.) for the Charlotte Harbor and Tampa Bay estuaries, respectively. These studies were developed and funded by the SWFWMD to identify the presence or absence of human pathogens and to develop Kurz, Harwood, Rose, and Lim Seventh Biennial Stormwater Research & Watershed Management Conference May 22-23, 2002 63 better indicators of microbial pollution. Both studies evaluated streamflow and rainfall conditions as predictors of microbial indicator concentrations. Both studies found significant positive relationships between hydrologic conditions and indicator concentrations. An analysis of data from EPC 143 did not, however, indicate a strong relationship between flow and fecal coliform concentrations (Figure 3a) in the case study area. Figure 1. Fecal coliform concentrations in the Hillsborough River watershed between 1995 and 1999 (HCEPC data). (Map source: Ayres Associates, Hillsborough River Watershed Management Plan, 2001) A similar analyses of fecal coliform concentrations at EPC 143 and rainfall from the nearby Plant City rain gauge did show a weak but significant positive relationship between these two parameters (Figure 3b). The consistently elevated concentrations found at EPC 143 indicates that one or more persistent sources of fecal contamination may exist in the region such as poorly constructed on-site wastewater treatment systems (OWTS), free-ranging livestock, or failing package plant wastewater treatment systems. The positive relationship with rainfall indicates that stormwater runoff facilitates transport of fecal material downstream from the various sources in the watershed. Kurz, Harwood, Rose, and Lim Seventh Biennial Stormwater Research & Watershed Management Conference May 22-23, 2002 64 Figure 2. Trends in monthly fecal coliform concentrations at EPC stations 108 and 143 between January 1988 and June 2000. Twelve (12) month moving averages are shown as solid colored lines for each station. Rainfall depths are for the Plant City gauge. (Source: Kurz and Harwood, Upper Hillsborough River Bacterial Contamination Assessment Plan of Study, 2001) Figures 3a and 3b. Scatterplots and regression analyses between fecal coliform concentrations and flow (left) and rainfall (right) for EPC 143 in the Blackwater Creek region. (Source: Kurz and Harwood, Upper Hillsborough River Bacterial Contamination Assessment Plan of Study, 2001) Upper Hillsborough River Fecal Coliform Trends (EPC Stations) 1 10 10