Skip to main content
ARS Home » Southeast Area » Athens, Georgia » U.S. National Poultry Research Center » Egg and Poultry Production Safety Research Unit » Research » Publications at this Location » Publication #421129
Research Project: Reduction of Foodborne Pathogens and Antimicrobial Resistance in Poultry Production Environments

Location: Egg and Poultry Production Safety Research Unit

Title: Random forest modeling to identify key Farm to Fork factors influencing Campylobacter level in Pastured Poultry systems

Author
item KIM, MINHO - Department Of Energy
item AL HAKEEM, WALID - Department Of Energy
item Rothrock Jr, Michael

Submitted to: International Poultry Scientific Forum
Publication Type: Abstract Only
Publication Acceptance Date: 11/25/2024
Publication Date: 1/5/2025
Citation: Kim, M., Al Hakeem, W., Rothrock Jr, M.J. 2025. Random forest modeling to identify key Farm to Fork factors influencing Campylobacter level in Pastured Poultry systems. International Poultry Scientific Forum. p. 322.

Interpretive Summary:

Technical Abstract: Campylobacter in poultry flocks poses significant food safety challenges, yet risk factors across different production systems remain unclear. Specifically, the impact of farming and environmental factors on Campylobacter dynamics in poultry production is not well established. This study aims to address these knowledge gaps by using random forest modeling to identify and rank the importance of farming practices, soil characteristics, and meteorological factors influencing Campylobacter levels in pastured poultry systems. Data were collected from 11 pastured poultry farms in southeastern United States from 2014 to 2017. Two separate random forest models were developed using different sets of predictor variables: (1) 32 farming practices and 26 soil physiochemical constituents, and (2) 80 meteorological variables. The models were applied to Campylobacter levels in five sample types: soil, feces, ceca, whole carcass rinse after processing (WCR-P), and whole carcass rinse after chilling and storage (WCR-F). Results from the first model identified farm, flock age, and season of sample collection as key predictors of Campylobacter levels. Farms A and B consistently showed higher levels in soil, feces, and ceca samples. However, WCR-F samples revealed higher levels in Farms E and J, indicating that preharvest conditions do not necessarily correlate with postharvest samples due to processing interventions. For WCR-P samples, the use of organic acids during processing was identified as the most effective intervention. The second model, focusing on meteorological factors, revealed that humidity-related variables, including maximum humidity two days prior to sampling and average minimum humidity six days prior to sampling, were two most important predictors for soil samples. For feces samples, the daily temperature range three days prior to sampling and average precipitation seven days prior to sampling were the most influential factors. This study provides valuable insights into the complex dynamics of Campylobacter in pastured poultry systems, identifying key predictors. Future research should focus on developing targeted control strategies that account for identified key factors.