Reducing the risk for transfer of zoonotic foodborne pathogens from domestic and wild animal to vegetable crops in the Southwest desert.

We propose to conduct a study that will help the leafy greens produce industry identify domestic and wildlife animal reservoirs of shiga toxin-producing E. coli (STEC) and Salmonella in the desert southwest growing region (DSGR) of the U.S. and northern Mexico, a region second only to the central California coast in production of lettuce. Additionally, we will identify management and risk mitigation practices that reduce contamination of leafy greens by these species. Scientific collections and epidemiologic methods will be used to complete three main objectives: 1) determine if terrestrial and avian wildlife species reported by the local growing community to most frequently intrude upon produce fields in the DSGR, are reservoirs of STEC and Salmonella; 2) determine the extent to which wildlife and livestock share genetically related strains of STEC and Salmonella, and measure the movement of strains from livestock operations to produce fields by wildlife populations; 3) extend knowledge of produce contamination prevention gained from the first two objectives to growers. Data from this collaborative study between industry, game management, and academic organizations will fill gaps in knowledge pertaining to animal intrusions and establishing buffer zones between adjacent livestock and produce operations in the desert environment.

Foodborne disease illnesses caused by pathogenic bacterial contamination of fresh produce are being recognized in greater numbers in the United States (Sivapalasingam et al, 2004; Lynch et al, 2009). The role of domestic and wildlife animals as reservoirs and transmitters of bacterial pathogens to fresh produce has been studied at length in some growing regions, such as the California central coast (Cooley et al, 2007; Jay et al, 2007; Gorski et al, 2011). In contrast, limited information exists on the importance of animal sources of contamination in the desert southwest growing region (DSGR), a major fresh vegetable production region of the US-Mexico border that is second only to central California in terms of lettuce production. These knowledge deficits may lead to uninformed environmental assessments and management decisions, especially those pertaining to mitigation efforts following animal intrusion and implementation of appropriate buffer zones between neighboring livestock and produce lands. The long-term goal of this project is to identify potential domestic and wild animal reservoirs of foodborne pathogens in leafy green production regions of the DSGR, and to use this knowledge to guide industry practices while minimizing impacts on wildlife species and their habitat. During the 2010-2011 DSGR season, our research team conducted the first phase of this study by examining local canid (dog, coyote) reservoirs in collaboration with leafy greens industry partners (Jay-Russell, 2011). This proposal represents the next phase of the research and will complete the following objectives: Objective 1: To determine if domestic animals and terrestrial and avian wildlife species in the desert southwest produce production region are reservoirs of shiga toxin-producing E. coli (STEC) or Salmonella. We will utilize livestock surveys, and hunter-harvested and targeted sampling and testing of feces and colon from terrestrial and avian wildlife species to assess the potential pathogen burden of these animals. Objective 2: To determine the extent to which wildlife and cattle share genetically related strains of STEC and Salmonella, and measure the movement of strains from livestock operations to produce fields by wildlife populations. We will utilize intensive trapping and radio telemetry with spatial analysis to assess animal movements and the potential spread of pathogens including evaluation of the interim Leafy Green Marketing Agreement guidance distance of 400 ft from a concentrated animal feeding operation to the edge of a crop. Molecular genotyping will provide a comparison of the genetic relatedness of strains isolated from these animals with strains from nearby livestock. Objective 3: Extend knowledge of preventing produce contamination by domestic animals and terrestrial and avian wildlife populations to the produce and livestock communities. We will share the knowledge gained from this study with growers, ranchers, buyers, regulators, conservation groups, and other stakeholders to improve best practices relating to pre-season and pre-harvest environmental assessments and wildlife intrusion, and no-harvest buffer zones. These objectives will be met by utilizing partnerships with local industry collaborators and the Arizona Game and Fish Department’s Research Branch. Hypotheses will be tested through bacteriologic and molecular microbiologic approaches, and spatial and statistical epidemiologic analysis. Results from this study will improve environmental assessments and intervention management decisions in support of both food safety and environmental stewardship.

1) To determine if domestic animals and terrestrial and avian wildlife species in the desert southwest produce production region are reservoirs of shiga toxin producing E. coli (STEC) or Salmonella. 

2) To determine the extent to which wildlife and cattle share genetically related strains of STEC and Salmonella, and measure the movement of strains from livestock operations to produce fields by wildlife populations. 

3) Extend knowledge of preventing produce contamination by domestic animals and terrestrial and avian wildlife populations to the produce and livestock communities.

• Desert fauna including wild rodents, birds, and javalina may serve as reservoirs of Salmonella and non-O157 STEC, but they do not appear to be significant sources of E. coli O157, even when in close proximity to high density cattle operations with documented fecal shedding of E. coli O157:H7. As shown in previous studies, cattle and feral swine fecal material is much more likely to contain E. coli O157:H7 compared with small mammals and avian species. However, due to the heat, feral swine are not found in the southern-most leafy green production areas of Arizona and California. 

• Wild rodent home ranges are small enough that these animals are often captured in the same location on multiple dates. It is unlikely that rodents travel from CAFOs to produce fields if the distance between the two is further than a 0.5 hectare area or if there are substantial barriers between the two, such as a canal or major roadway. Therefore, it is highly unlikely that wild rodents transmit pathogens directly from CAFOs to produce fields. 

• Rodent density was relatively low at the CAFOs, but high (>30% trap success) along leafy green field edges and adjacent habitat; additional studies should evaluate if the buffer (typically dirt roads between fields) between rodent habitat and produce is sufficient to minimize intrusions into the fields. 

• Birds travelled regularly between CAFOs and distant sites, sometimes on a daily basis, and their flight paths sometimes take them directly over produce fields. Since birds do show seasonal variation in their prevalence levels of Salmonella and non-O157 STEC, growers are advised to deter birds from their fields during the growing and harvesting season. 

• Bird density and diversity was highest at the CAFO and in natural habitat adjacent to produce fields. During the study, birds were rarely captured within fields suggesting that they spent most of their time between CAFOs and roosting areas. Future studies should focus on the specific attractants that cause birds to enter produce fields (e.g., irrigation, harvest, etc.) and target bird control resources to these higher risk activities. 

• If leafy green fields are near CAFOs and experience significant bird intrusions (especially large groups of brown-headed cowbirds, ravens, grackles, blackbirds, and sparrows), the minimum 400 foot buffer zone between the field and the perimeter of a CAFO, recommended as an LGMA food safety practice, may need to be increased. 

• The majority of STEC isolates obtained in this study were not among the “top 6” non-O157 serogroups (O26, O111, O103, O121, O45 and O145) and did not encode genes for intimin and other virulence markers typically associated with human infections. Rapid discrimination of clinical relevance would improve risk management decisions and, potentially, avoid unnecessary destruction of crops.