P1-164 Decline of Salmonella on Artificially Contaminated Dry Pet Food

Monday, August 4, 2014
Exhibit Hall D (Indiana Convention Center)
Elisabetta Lambertini, University of Maryland, College Park, MD
Huilin Cao, University of Maryland, College Park, MD
Miao Guo, University of Maryland, College Park, MD
Abhinav Mishra, University of Maryland, College Park, MD
Robert Buchanan, University of Maryland, College Park, MD
Abani Pradhan, University of Maryland, College Park, MD
Introduction: Due to several outbreaks and large-scale product recalls, Salmonella has emerged as a priority pathogen in dry pet food and treats. However, little data is available to quantify the risk posed by this class of products to both pets and their owners. Specifically, the kinetics of Salmonella survival on this complex food matrix composed of fats, protein, and grains has not been characterized.

Purpose:  This study sought to measure the short- and long-term kinetics of survival or decline of Salmonella in dry pet food under storage conditions (low water activity and ambient temperature) commonly encountered during production, at retail, and at households.

Methods:  A Salmonella enterica cocktail of 12 strains isolated from pet food and treats was used to inoculate 25-g batches of commercial dry pet food at approximately 9.5 log CFU/sample. After air drying for 8 hours, inoculated samples were stored in airtight containers at room temperature (21-23°C). At each sampling time, Salmonella was enumerated in three inoculated and one negative sample by stomaching in 0.1% peptone water, preparing serial dilutions, and plating on generic (BHI) and selective (XLD and BSA) media using a spiral plater. Water activity was also measured. Resulting data points were fitted using log-linear and non-linear decline models using the R software.

Results: Results at 83 days highlight a fast and approximately log-linear initial decline (up to 54 days), followed by a phase of lower decline rate. A Weibull model (Log CFU ~ Log CFU0 - (1/delta)tp) provided a satisfactory overall fit (delta: mean 6.91, SE 2.76 log CFU/days; p: mean 0.48 Log CFU/days, SE 0.07). Water activity remained constant at 0.48-0.50.

Significance: This study provides a complete decline model that can feed into quantitative risk assessments.