Purpose: This study aimed to compare different transfer modes of cross-contamination associated with Salmonella during almond shelling (shell to kernel) and sorting (kernel to kernel).
Methods: Almond shell pieces and kernels (200 g) were inoculated with Salmonella Enteritidis PT30. The inoculated kernels (5 g), shell pieces (5 g) and uninoculated kernels (200 g) were conditioned at 0.40 aw. Uninoculated kernels and either inoculated shell pieces or kernels were then mixed in a stainless steel drum (140 mm diameter, 64 mm depth) for the total number of rotations (TNR) of 5, 20, 40, and 80 revolutions (rotational speeds of 8, 16, and 24 rpm for 10 to 300 s), in triplicate. At each condition, uninoculated samples (four kernels) were retrieved, stomached, and enumerated for Salmonella on modified trypticase soy agar.
Results: Initial inoculation levels on the kernels (8.37±0.50 log CFU/g) and shell (5.94±0.04 log CFU/g) were significantly different (P<0.05). The maximum bacterial load transferred from shells to kernels (2.36±0.26 log CFU/g ) was lower (P<0.05) than that of kernels to kernels (4.53±0.41 log CFU/g). However, there was no significant difference in shell inoculation level (5.55±0.41 log CFU/g, P=0.13) before and after the transfer experiment, which indicates that most bacteria remained on the shell.
Significance: Understanding bacterial cross-contamination modes and mechanisms is critical information for secondary modeling, which will enable elucidation of the cross-contamination pathways of Salmonella for an entire processing system.