Growth dynamics and predictive modeling of Salmonella Enteritidis isolated from the commercial broiler farm environment

Authors: FATIMA, ARJMAND - Auburn University; NAEEM, MUHAMMAD - Auburn University; BAILEY, MATTHEW - Auburn University; BOURASSA, DIANNA - Auburn University

Submitted to: Journal of Applied Poultry Research
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 12/22/2025
Publication Date: 12/23/2025
Citation: Fatima, A., Naeem, M., Bailey, M., Bourassa, D. 2025. Growth dynamics and predictive modeling of Salmonella Enteritidis isolated from the commercial broiler farm environment. Journal of Applied Poultry Research. https://doi.org/10.1016/j.japr.2025.100661.
DOI: https://doi.org/10.1016/j.japr.2025.100661

Interpretive Summary: Salmonella Enteritidis is a leading cause of foodborne illness globally, particularly associated with poultry products. Rapid and accurate prediction of bacterial growth in poultry environments is essential for effective contamination control. This study aimed to assess growth dynamics and develop a predictive linear regression model to estimate the growth of nalidixic acid-resistant Salmonella Enteritidis isolated from a broiler farm exhaust fan, using optical density (OD) and incubation time as predictors. In-vitro experiments were conducted over 18 h, with hourly measurements of viable cell counts (CFU/mL) and OD, capturing the full bacterial growth cycle. Stepwise regression was used to construct a model based on log (ln)-transformed CFU, OD, and time data, which achieved high accuracy and demonstrated reliable cross-validation performance. The final model equation integrated OD and its interaction with time, providing a practical method for estimating bacterial load without the need for labor-intensive plating. Results showed that OD alone does not adequately reflect viable counts unless contextualized with incubation time. This modeling approach addresses a key gap in the literature by focusing specifically on Salmonella Enteritidis from poultry-associated matrices rather than generalized serotypes or reference strains. The findings support the use of OD-based prediction models for real-time microbial monitoring in poultry processing and production settings. The model has potential applications in assessing contamination risk, facilitating rapid decision-making, and developing automated detection systems for food safety. Future work may extend this model to other environmental samples from broiler farms and serotypes to improve its broader applicability in poultry industry microbiology.

Technical Abstract: Salmonella enterica subsp. enterica serovar Enteritidis (Salmonella Enteritidis) is a leading cause of foodborne illness globally, particularly associated with poultry products. Rapid and accurate prediction of bacterial growth in poultry environments is essential for effective contamination control. This study aimed to assess growth dynamics and develop a predictive linear regression model to estimate the growth of nalidixic acid-resistant Salmonella Enteritidis isolated from the broiler farm exhaust fan, using optical density (OD) and incubation time as predictors. In-vitro experiments were conducted over 18 h, with hourly measurements of viable cell counts (CFU/mL) and OD, capturing the full bacterial growth cycle. Stepwise regression was used to construct a model based on log (ln)-transformed CFU, OD, and time data, which achieved high accuracy (R² = 0.910, RMSE = 0.591) and demonstrated reliable cross-validation performance. The final model equation integrated OD and its interaction with time, providing a practical method for estimating bacterial load without the need for labor-intensive plating. Results showed that OD alone does not adequately reflect viable counts unless contextualized with incubation time. This modeling approach addresses a key gap in the literature by focusing specifically on Salmonella Enteritidis from poultry-associated matrices rather than generalized serotypes or reference strains. The findings support the use of OD-based prediction models for real-time microbial monitoring in poultry processing and production settings. The model has potential applications in assessing contamination risk, facilitating rapid decision-making, and developing automated detection systems for food safety. Future work may extend this model to other environmental samples from broiler farms and serotypes to improve its broader applicability in poultry industry microbiology.