Bacteriological raw-milk quality: Factors influencing the relationship between colony-forming units and Bactoscan-FC counts

The proprietary system Bactoscan-FC is registered as a routine method for testing the bacteriological quality of raw milk on the precondition that, before grading, the measured values are converted into colony- forming units of the official/reference method. Due to the very different principles of measurement, it is necessary that the validity of the applied conversion be checked regularly. In the framework of such a check (1999-2002), and in addition to the respective measurements by both methods from raw- milk samples from dairy farms in Northern Germany, the following aspects of production were recorded: delivery interval, herd size, breed, milk storage, cooling, season and year. Regression analysis showed that the relationship between both methods comes so close to the conversion characteristic fixed for Germany in 1998/1999 that respective updates currently appear not to be necessary. An analysis of variance was computed in order to uncover the effects of special situations in production: the simple regression formula between methods without correction factors explained 67% of the variance (r 2 ). The explained variance increased by only 1 % if the "season" factor was included. By the inclusion of all influence factors considered in this study, the share of the explained variance increased only by 3% to 70%.