Influence of sampling intervals on the standard plate counts of milk samples

Document Type


Publication Date



American Dairy Science Association


Journal of Dairy Science




Suppl. 1






sampling, variability, intervals


High accuracy in sampling is critical in determining the microbiological quality and safety of milk and dairy products. The issue arises with the lack of a standard approach to lower the variability of sampling results of low count milk (LCM) and high-count milk (HCM). It may occur in extended and multi-processing conditions that are triggered by the heterogeneous and aggregating nature of bacteria. This study focuses on the influence of sampling intervals on the microbial count variability of LCM and HCM samples. For a pilot-scale study, 380 gallons of raw whole milk (<4°C) were spiked with Bacillus licheniformis vegetativecells (4 log cfu/mL) and pasteurized at 72°C/16 s at a 0.5 gpm flow rate for 12 h. Samples (100mL each) were drawn aseptically at intervals 0, 4, 8, and 12 h from each of the raw (HCM) and pasteurized (LCM) sides of the unit operation using commercially available a port-septum- based sampling device with needle insertion. Two continuous composite samples (2-L each) were also collected during the 12 h run in sterilized sampling bags at 2.5mL/min rate and held at < 4°C. Samples were analyzed at each interval for standard plate counts using tryptic soy agar and incubated at 32°C for 48 h. The study was done in duplicates and ANOVA was used to compare the microbial counts. Mean counts for the individual samples at 0, 4, 8, 12 h., and the composite samples at 12 h., were found to be 4.18 ± 0.10, 4.29 ± 0.12, 4.24 ± 0.12, 4.41 ± 0.12, and 4.38 ± 0.13 log cfu/mL for the HCM, and 1.52 ± 0.13, 1.57 ± 0.13, 1.66 ± 0.12, 1.69 ± 0.12, and 1.78 ± 0.10 log cfu/mL for the LCM, respectively. Results showed an increasing trend in the counts for LCM with no significant difference between the intervals (P > 0.1). The microbial counts of samples from various intervals compared with composite samples were also not significantly different (P > 0.1) for both LCM and HCM. Under the conditions of analysis, a single sample of 100 mL volumes at any interval during processing is as effective as composite sampling for estimating the microbial quality of milk under prolonged dairy processing conditions of 12 h.

This document is currently not available here.