Table 2
Correlation between different raw milk tests and various measures of pasteurized milk quality
Raw milk
test1 Correlation (R2)2
d 17
SPC d 17 SPC
(w/o PPC) d 21
SPC d 21 SPC
(w/o PPC) d 17 Sensory
score d 17 Sensory score
(w/o PPC) d 17
LP d 21
LP
SCC 0.0221 0.0005 0.0235 0.0016 0.1132 0.1571* 0.0014 0.0092
SP 0.0031 0.0001 0.0000 0.0058 0.0301 0.0223 0.0002 0.0140
Edwards 0.0092 0.2637 0.0002 0.2186 0.0283 0.0212 0.0109 0.0180
VJ 0.0011 0.0950 0.0003 0.1274 0.0145 0.0229 0.0569 0.0266
LP 0.0299 0.1009 0.0365 0.1122 0.0004 0.0060 0.0882 0.0509
Raw SPC 0.0544 0.2095 0.0463 0.1877 0.0000 0.0000 0.0896 0.0523
PBC 0.1280 0.2173 0.0876 0.1228 0.0097 0.0038 0.0601 0.0798*
CVTA 0.1238 0.4169* 0.1125 0.3681* 0.0003 0.0092 0.0907* 0.0678
Coliform 0.1301 0.2641 0.1471 0.2413 0.0565 0.0461 0.0399 0.0064
PI 0.2416* 0.1973 0.2211* 0.1874 0.1060 0.0725 0.0003 0.0003
ΔPI 0.1807 0.1524 0.1554 0.1291 0.1314* 0.0628 0.0436 0.0270
1SP = spore pasteurization; Edwards = growth on Edwards medium for enumeration of Streptococcus spp.; VJ = growth on Vogel Johnson medium for enumeration of Staphylococcus spp.; LP = laboratory pasteurization at d 0; PBC = psychrotrophic bacteria count; CVTA = growth on crystal violet tetrazolium agar for enumeration of gram-negative organisms; PI = preliminary incubation; ΔPI = change in bacterial numbers from raw SPC to PI.
2R2 values were obtained from linear regression models of log-transformed raw milk test data and various measures of pasteurized milk quality; R2 values marked with a * indicate the highest R2 values for a raw milk test associated with each pasteurized milk quality parameter. Pasteurized milk quality parameters included SPC at d 17 and d 21, average sensory score at d 17 and 21 as well as LP counts at d 17 and 21. For pasteurized milk SPC and sensory scores, R2 values were calculated for all samples or only those samples that showed no evidence for post-pasteurization contamination (PPC).
Thumbnail image of Figure 1. Opens large image
Figure 1
Scatterplot matrix of log-transformed data depicting the relationship between results from raw milk tests and various measures of pasteurized milk quality (data points represent relationships for 43 raw milk and 43 pasteurized milk samples collected from 4 New York State fluid milk processors). Ellipses encompass 95% of data points. PI = preliminary incubation; VJ = growth on Vogel Johnson medium for enumeration of Staphylococcus spp.; CVTA = growth on crystal violet tetrazolium agar for enumeration of gram-negative organisms; Edwards = growth on Edwards medium for enumeration of Streptococcus spp.; LP = laboratory pasteurization; PBC = psychrotrophic bacteria count; ΔPI = change in bacterial numbers from raw SPC to PI; LPC = laboratory pasteurization count; PPC = post-pasteurization contamination.
View Large Image | View Hi-Res Image | Download PowerPoint Slide
Thumbnail image of Figure 2. Opens large image
Figure 2
Linear regression analysis of (A) log-transformed d 17 SPC counts of 2% pasteurized milk versus log-transformed raw milk preliminary incubation (PI) counts; (B) d 17 sensory scores of 2% pasteurized milk versus log-transformed raw milk PI counts; and (C) log-transformed laboratory pasteurization count (LPC) for d 17 and d 21 versus log-transformed raw milk PI counts. Each point represents data from 1 of 43 samples collected from 4 New York State fluid milk processors. For panels A and B, data for sample pairs where the pasteurized milk sample showed evidence for post-pasteurization contamination (PPC) are shown as black diamonds, whereas sample pairs in which the pasteurized milk sample showed no evidence for PPC are shown as gray squares. In panel B, 4 of the 5 sensory data points represented by black diamonds (1 sample each from plants B and C and 2 samples from plant D) that fell below a score of 6.0 showed evidence for PPC, whereas 1 sample (represented by the gray square, sample from plant D) was characterized by predominance of Bacillus spp. and Paenibacillus spp. and showed no signs of PPC. Linear regression lines for all samples and for only those samples that show no evidence for PPC are also shown in panels A and B.