Summary of Awards to Date

Remotely-sensed and field-collected hydrological, landscape and weather data can predict the quality of surface water used for produce production


Jan. 1, 2017 - Dec. 31, 2018

Award Number


Amount Awarded



Martin Wiedmann, Ph.D
Cornell University


Channah Rock, Ph.D.


There is a clear need for the development of improved, science-based tools to help reduce pre- harvest introduction of microbial produce safety risks through surface water use. The purpose of this project is (i) to identify and prioritize spatial and temporal risk factors for microbial contamination of surface water, and (ii) to develop geospatial models that predict surface water microbial quality, which will be assessed by quantifying generic E. coli and testing for key pathogens (e.g., Salmonella). Spatial and temporal variation in water quality will be assessed by repeatedly testing multiple water sources over two years. Publicly available remotely sensed data (e.g., predominant upstream land-use) will be used to identify factors that are associated with elevated E. coli levels, and an increased risk of pathogen detection. Data collection will be performed in two produce growing regions (AZ and NY) to assess the robustness of our models and their translatability to other regions. These data and models will allow growers to identify times and locations where surface water sources are more likely to be microbially contaminated. This will enable growers to better time water use, testing, and treatment to minimize produce safety risks associated with microbially contaminated surface water.

Technical Abstract

The quality of surface water used for produce production (e.g., irrigation, frost protection) has emerged as a key issue for preventing pre-harvest microbial contamination of produce. As part of the Food Safety Modernization Act (FSMA), the FDA established microbial standards for the use of surface water for produce production, including how frequently water should be tested. However, surface water quality is known to vary based on adjacent or upstream land- use and weather (i.e., meteorological) factors. Therefore, targeted approaches to water testing

and treatment that account for this variation may improve growers’ ability to identify and address on-farm food safety risks associated with surface water use. Due to the widespread availability of remotely sensed (RS) data, statistical analyses (e.g., predictive models) that utilize RS data can be used to develop such targeted approaches for individual water sources and farms. This project is designed to develop such approaches for surface water in two produce-growing regions, AZ, and NY; water quality will be assessed by quantifying the concentration of generic Escherichia coli, and testing for the presence of key pathogens (Shiga-toxin producing E. coli [STEC], Salmonella, and Listeria monocytogenes) in surface water samples. The end product of our project will be geospatial models that can be used to predict surface water quality for individual water sources. These models will serve as a framework that can be built upon as more data becomes available. As part of model development, consistent and region-specific risk factors for microbial contamination of on-farm surface water will be identified and prioritized.

This prioritized list will provide growers with information that they can use to generally predict times and locations with an increased risk of pathogen detection and/or elevated E. coli counts. As the list and models will account for spatial and temporal factors, growers will be able to adjust surface water use and target risk management efforts for specific times and water sources. Additionally, since we will collect data on generic E. coli levels and the presence of STEC, Salmonella, and L. monocytogenes, we will quantitatively assess, in different regions, the association between E. coli concentration and pathogen presence in surface water used for produce production, and how this association is affected by changes in spatial and temporal factors. For example, E. coli levels and pathogen detection may be correlated if certain conditions are met (e.g., after rain and when the predominant upstream land-use is pasture), but not under other conditions. This type of information will help clarify some of the contradictory data on the association between E. coli levels and pathogen detection in on-farm surface water (4, 19, 31) and will provide new data that can help with the assessment of the pathogen presence risk associated with higher E. coli levels.  Overall, this project will increase our understanding of pre-harvest produce safety risks that are associated with surface water use, which will allow growers and the produce industry to develop strategies to optimize, at the individual farm level, produce safety efforts.