COMPARISON OF METHODS FOR GRADING MILK INTENDED FOR MANUFACTURING PURPOSES1,2

Abstract

Summary Samples of milk intended for manufacturing purposes were obtained once each season from randomly selected dairy farms in three widely separated geographical locations. Samples (3873) were analyzed by standard plate count (SPC-32 C), direct microscopic clump count (DMCC), methylene blue (MBRT) and resazurin (RRT), (5P 7/4) test. Correlations for SPC vs. DMCC, MBRT, and RRT were r = 0.78, −0.82 and −0.79, respectively; DMCC vs. MBRT, RRT, r = −0.75 and −0.68, respectively; MBRT vs. RRT, r = 0.86. Prediction of DMCC, MBRT, and RRT on the basis of their regressions with SPC Were influenced significantly (99% level of significance) by the geographical source of samples, the type of milk handling on the farm (can or farm bulk tank) and season. On the basis of equivalents (determined by regression) between classifying tests, new standards were developed. The percentage of samples classed the same by the four methods using these new calculated classifications or the classification recommended by the United States Department of Agriculture (USDA) was approximately 50% of all samples. Agreement between the four methods of classifying samples was not close enough to warrant the interchangeable use of the four tests. Results from calculated classifications varied considerably with geographical location and with the method of handling milk on the farm. Although the use of calculated classifications would give better agreement among the four methods when applied to the supplies from which they were derived, multiplicity of tests methods, their application and interpretation would cause much confusion. Use of the RRT with a 5P 7/4 Munsell color end-point with separate procedures, one for can supplies (RRT: > 2 1/2, ≤ 2 1/2, and ≤ 1 1/2 hr for Class 1, 2, and 3, respectively) and one for farm bulk tank supplies (RRT: > 3 1/2, ≤ 3 1/2, and ≤ 2 1/2 for Class 1, 2, and 3, respectively) resulted in grading milk, intended for manufacturing purposes, more uniformly than either use of the four classification methods interchangeably or even use of a single RRT standard based on all samples regardless of type of milk handling on the farm (can or bulk tanks).