Published in: American Bar Association, ABA Water Quality and Wetlands Committee Newsletter, Fall 2010
MICROBIAL SOURCE TRACKING: POINTING THE FINGER OF BLAME FOR WATERBORNE PATHOGENS
GUY R. KNUDSEN
Clear, sparkling water is not an accurate indicator that rivers, lakes, streams, or private and municipal water sources are free of microbial contamination. The microbiological impairment of water quality, resulting from activities including agriculture, industry, and residential waste generation, is one of the most significant threats to human health. A variety of sources contribute potentially pathogenic microorganisms to waters, including fecal contaminants originating from humans, livestock, and wildlife. Counts of fecal coliform bacteria form an important component of TMDL (Total Maximum Daily Load) development and analysis. Under section 303(d) of the 1972 Clean Water Act, states are required to develop lists of impaired waters, and to establish priority rankings for waters on the lists and develop TMDLs for them. TMDLs specify the maximum amount of a pollutant that a waterbody can receive and still meet water quality standards, and allocates pollutant loadings among point and nonpoint sources. Fecal coliform counts are standard for residential well testing, and are commonly used in watershed monitoring by regulatory agencies. Many TMDL watersheds are impaired by excess fecal coliform levels. For example, it has been estimated that 47% of Minnesota’s river and stream miles do not achieve the Clean Water Act’s ‘swimmable’ goal due to elevated fecal coliform counts. Nationwide, the U.S. EPA estimates that about 13% of the country’s surface waters violate fecal bacteria standards.
The large group of bacteria known as ‘coliforms’ includes a number of genera, which thrive in various soil and aquatic environments. A subgroup, fecal coliform bacteria (especially, Escherichia coli) reside in the intestines of warm-blooded animals, and are excreted in waste material. The presence of E. coli is specifically indicative of fecal contamination, and high numbers of these bacteria in water samples traditionally has been used as an indicator of unsanitary conditions that pose health concerns. Thus, detection of fecal bacteria may also indicate the presence of toxigenic E. coli strains such as O157:H7, as well as other pathogenic bacteria (Shigella, Salmonella), enteroviruses such as Hepatitis A or Norwalk group viruses, and parasites such as Giardia or Cryptosporidium.
A certain level of fecal coliform bacteria will always be found in watersheds. Human sewage, concentrated animal feeding operations (CAFOs), livestock grazing, and agricultural operations such as manure application all contribute fecal contaminants. Waterfowl and other wildlife may also contribute significantly to bacterial counts. Knowing the source(s) of bacteria in a water body or water supply is potentially of great value in the remediation and prevention of further bacterial contamination, as well as in efforts to ascribe liability to those responsible for the pollution.
To date, there are no universally accepted standard methods to distinguish among human, livestock, wildlife, and other sources of fecal coliform pollution. However, the development of new methodologies for determining the actual animal source of fecal coliform bacteria (as well as certain other fecal microbes) has been an area of intensive research for a number of years. In a best-case scenario, waterborne E. coli can be characterized by genetic or phenotypic means, to the extent that it now becomes possible to pinpoint specifically where the bacteria came from (e.g., CAFO, streamside grazing, residential sewage, etc.), thus greatly expanding the potential for litigation against those responsible for the pollution. This suite of new methods is known by the generic name of ‘microbial source tracking’ (MST), or, more specifically for E. coli and other bacteria, as ‘bacterial source tracking’ (BST). In this article, I will focus on BST methods developed for E. coli, and the role these are expected to increasingly play in environmental regulation and litigation.
Each of the BST methods relies on the observation that different animal sources produce unique, identifiable strains of fecal bacteria, presumably as a result of the different intestinal environments and selective pressures to which the bacteria are subjected. A bacterial strain that has adapted to a particular host intestinal tract is capable of colonizing that environment and competing favorably with other members of the indigenous intestinal microbial community. The BST methods that have been investigated can be broadly categorized according to whether they distinguish among samples based on their genetic makeup (genotypic methods), or based on their observable physiological characteristics (phenotypic methods). Within each group, some methods require the development of a background ‘library’ or database, against which an unknown sample can be compared, while other methods do not require such a library. Genotypic methods generally involve analysis of specific DNA sequences from the sample organism, typically after amplification (via polymerase chain reaction, PCR) of the tiny amount of bacterial DNA that can be recovered from environmental samples. One particularly promising method, known as ribotyping, analyzes a portion of the organism’s genome that encodes for ribosomal RNA. Ribosomal RNA is the central component of ribosomes, the protein factories possessed by all living cells. Because the function of ribosomal RNA is so essential, the encoding genes tend to change very slowly over evolutionary time, allowing relatedness among species and even subspecies to be narrowly defined with a characteristic ‘fingerprint’. Phenotypic methods of bacterial source tracking evaluate such characteristics as resistance to (sometimes multiple) antibiotics, or profiles of specific carbon source utilization.
Each of the many BST methods has its proponents, but generally the genotypic methods have been found to more rigorously distinguish fecal bacteria from different sources, although typically at a higher cost in materials and labor. Conflicting criteria for determining an optimum method, i.e. accuracy and repeatability vs. low cost, will play a role in management decisions as well as potential litigation, where results will be subjected to evidentiary standards.
Although development of MST methods is ongoing, they are proving to be powerful tools for watershed management, and are increasingly being applied by public agencies. Particularly in California, agencies are spending millions of dollars to identify sources of bacteriological water contamination. For example, the city of San Diego commissioned two microbial tracking studies in 2002, after extensive sewer and storm drain upgrades failed to reduce bacterial pollution in Mission Bay. That study concluded that bird droppings were the primary source of the pollution, and the city was able to modify its bayside irrigation practices to alleviate the problem. When surfers at Rincon Point, near Santa Barbara, complained about bacterial pollution in the lagoon, the environmental group Heal the Ocean teamed with county health officials to commission a BST study. There, the source was tracked to septic tanks from neighboring homes, prompting a neighborhood upgrade of the sewer system. On another beach, however, a BST study prompted a different management strategy. At Arroyo Quemado Creek beach, frequently closed due to high bacterial counts, the study indicated that seagull droppings were the primary contribution to the problem, and when the birds were scared away the beach water quality rose dramatically.
The development of BST methods has implications for CAFOs as well. There is increasing public concern about potential water pollution from livestock feeding operations. For example, the Texas Institute for Applied Environmental Research found that elevated fecal coliform levels in rivers were correlated with the application of dairy manure to nearby fields. EPA has proposed strict Animal Feeding Operations guidelines in an attempt to prevent manure pollution from those operations getting into waterways and other riparian areas. As pointed out by the environmental management committee of the National Cattlemen’s Beef Association, hopefully “regulations will be based on a cooperative basis and not on a top-down basis”, and “a site-specific, science-based approach is the only way to make environmental regulations concerning livestock production operations work.” With growing public concern about the relationship between livestock operations and water pollution, more citizen complaints can be expected, and potential consequences for the industry will be stricter regulations, higher costs, penalties, and even lost business.
Of course, identification of a source of fecal bacterial contamination can have implications for more than management decisions. There is a long history of successful lawsuits by plaintiffs whose health was adversely affected by ingestion of food or water contaminated by fecal bacteria. The most notorious of these have been related to E. coli contamination of food items, including hamburgers, strawberries, and spinach. In the case of food crops, it is likely that the contaminant bacteria contacted the produce via polluted irrigation water, although the chain of causation has typically not been easy to demonstrate. Application of BST methods should make the assignment of liability more feasible, and has the potential to greatly increase litigation in this arena.
However, one recent case indicates that the courts may not yet be ready to accept BST methods, at least in the current state of the art. In May of this year, the U.S. Tenth Circuit Court of Appeals affirmed a district court decision not to enjoin Tyson Foods, Inc. and several other defendants from applying poultry waste to lands within the Illinois River watershed (Oklahoma v. Tyson, 10th Cir. 2009, No. 08-5154.) Poultry litter contains a number of enteric bacterial pathogens including E. coli, Salmonella, and Campylobacter. To bolster its claim that land application of poultry litter is a cause of fecal bacterial contamination in the Illinois River waterways, the state presented two expert witnesses, a microbiologist and a geochemist, who testified that their use of BST traced a contaminant path from fields where the waste was applied to the watershed, with detection of a ‘poultry-specific’ signature. The district court rejected that contention, however, concluding that Oklahoma failed to demonstrate that “bacteria in the waters of the IRW are caused by the application of poultry litter rather than by other sources, including cattle manure and human septic systems.” (Oklahoma v. Tyson, D.C. No. 05-CV-00329-GKF-SAJ (N.D. Okla.)) The district court further held that the expert testimony was not sufficiently reliable under the Daubert standards (Daubert v. Merrell Dow Pharmaceuticals, 509 U.S. 579 (1993)), and therefore declined to accord it the weight required to establish a link between the poultry litter and the fecal bacteria levels in the rivers and streams. The appellate court agreed and affirmed, adding that in its opinion the BST technology employed was “novel and untested.”
Although the Oklahoma v. Tyson court appeared to support the contention of Tyson’s attorney that the state was relying on claims of “ghost pathogens” to make its case, an increasing volume of scientific evidence and acceptance of MST technology will likely serve to limit the precedential value of this decision. Source tracking of microbial pollutants will undoubtedly play a major role in future litigation in the water quality arena.
Guy R. Knudsen is Professor of Microbial Ecology & Plant Pathology at the University of Idaho, and a private practice attorney in environmental law. You may contact him via e-mail at firstname.lastname@example.org.