Summary of Awards to Date

Wildlife survey for E. coli O157:H7 in the central coastal counties of California


Oct. 1, 2009 - Nov. 30, 2011

Award Number


Amount Awarded



Andrew Gordus
California Department of Fish and Game


E. Robert Atwill, Ph.D., Robert Mandrell, Ph.D.


We propose to continue testing wildlife for the human pathogenic strain of bacteria, Escherichia coli O157:H7 in Monterey, San Benito and San Luis Obispo Counties. Since 1995, this pathogen has resulted in more than 25 outbreaks from eating leafy green vegetables (LGV); approximately half of these have been associated with LGV grown on the California central coast. Wildlife has been suggested to be a source of E. coli contamination of LGV. Because of this uncertainty, farmers are being required to build deer- and wild pig-proof fences around their fields, and remove habitat and wildlife from their farms. To date, however, there is minimal definitive data that wildlife are an important source of contamination.
We propose to collect colon or fresh fecal samples from wildlife collected in relevant LGV production areas to determine if they are carrying E. coli O157:H7. This information will help us better manage and protect wildlife and provide food health safety information to farmers and to the food industry. The future of sustainable wildlife populations in the three central coastal counties is dependent on having cumulative and accurate scientific data to properly manage wildlife and to protect human health.
Technical Abstract
In 2006, human pathogenic strains of Escherichia coli O157:H7 (EcO157) were isolated from non-native wild pigs in San Benito County. EcO157 can cause serious health problems in people. However, to date, minimal data exist to determine if wildlife are a true source of contamination. We propose to collect samples to determine if wild animals are carrying EcO157. With better knowledge about the spatial and temporal incidence of EcO157 in local wildlife, we will provide accurate data for growers, landowners, processors and auditors in order for them to make decisions that will balance food safety concerns with wildlife management. The results will assist resource agencies and growers in developing strategies, and management plans and policies for preventing crop contamination in the fields to protect public health and to protect wildlife and their habitats. The purpose is to bring practical science and farm practice to the farming and ranching communities. This will help us better manage and protect wildlife and for agencies to develop policy and wildlife management plans to reduce the food safety wildlife uncertainty.
The coastal county area has been divided into watersheds using Department of Water Resources watershed maps. The reasoning for a watershed-based landscape is that EcO157 is believed to wash down watersheds, most animals will have lived most of their life within a particular watershed, and land ownership does not need to be identified.
We propose to collect colons or anal swabs from birds, small mammals, wild pigs, and deer. If provided the opportunity by landowners, we will collect samples from elk, geese, and ducks; some samples will be obtained at public sites. Small birds will be captured using mist nets and small mammals will be captured using live-box traps and released. We plan to coordinate with landowners or with USDA - Wildlife Services to collect colons from wild pigs during depredation hunts and trapping. All samples will be analyzed by USDA-ARS-WRRC or UC Davis – Western Institute for Food Safety and Security (WIFFS) laboratories by methods reported previously by Cooley 2007 and Jay 2007. The results will be forwarded to Dr. Rob Atwill at UC Davis for further statistical analysis and publication.
Because of the sensitivity of this issue, landowners do not want to be identified. As such, UC Davis has developed a sampling location coding system, so land ownership can not be identified by laboratory staff and the results are not available to field staff. The analyzed results will be published based on a large geographical scale, such as watersheds or by county, thus preventing identifying individual land ownership. We are networking with both public and private landowners to seek voluntary permission to allow us to collect samples discreetly and confidentially the next 2 years. To date, we have access permission to over 14 properties, plus we continue to make contact with other land owners.