A range of bacterial levels was observed among samples within check details a single time point, sometimes resulting in a standard deviation that was larger than the average counts, in part, a product of the combination of enumerated and assigned values for samples ( Table 2). This is an inherent limitation of microbiological

data. Gathering statistically sound plate count data is only possible when using higher inoculum concentrations, but such treatments are less likely to mimic natural contamination scenarios. Lowering the limit of detection for enumeration by filtration or including MPN determinations would add time and cost to the analysis but should be considered for future studies. Differences among samples in nut shell topography and shell integrity (e.g., small Galunisertib molecular weight cracks) may also have contributed to this variation by influencing our ability to remove inoculated organisms with our sampling procedure. Bacterial decline at each sampling point during storage

was calculated by subtracting the levels determined at the sampling point from the levels measured at the beginning of storage. The declines among the three genera were similar at all sampling points except for three; greater declines were observed in L. monocytogenes populations than in Salmonella and E. coli O157:H7 populations at 27, 83, and 97 days of storage. Over 97 days of ambient storage, declines of Salmonella and E. coli O157:H7 were estimated to be less than 1 log CFU/nut and the decline of L. monocytogenes was 2 log CFU/nut. These declines were less than the 2.8- to 3.8-log decline observed for Salmonella Enteritidis PT 30 on walnuts inoculated at 10 or 7.5 log CFU/nut, respectively, and stored for a similar length of time

(83 to 139 days) ( Table 1). These data are comparable to previous studies with other tree nuts; as populations decrease to near the standard LOD the rate of decline slows ( Beuchat and Heaton, Sclareol 1975, Beuchat and Mann, 2010a, Blessington et al., 2012 and Kimber et al., 2012). During the 97-day storage period, 78 inshell nuts were sampled per genera; of these, 73 Salmonella-, 66 E. coli O157:H7-, and 66 L. monocytogenes-inoculated nuts were positive by plate count or enrichment. Plate counts of at least 1 log CFU/nut were obtained for 49 Salmonella-, 23 E. coli O157:H7-, and 31 L. monocytogenes-inoculated nuts. At all time points during storage after the initial plating, all samples were subjected to a primary enrichment. Enriched broths were streaked onto selective/differential media for confirmation if enumerated values were below the LOD or if the previous enrichment was negative. An additional number of walnuts were positive after secondary or tertiary enrichment ( Table 2); 14 nut samples (6% of the 234 nut samples evaluated) required additional enrichment beyond the initial 24 h for positive isolation. Recovery of pathogens from dry foods presents a challenge as the cells may be severely injured.