Incidence and characterization of Bacillus cereus in two dairy processing plants

In order to determine the contamination of raw and pasteurized milk with (psychrotrophic) Bacillus cereus, samples were taken at two dairy processing plants, from raw milk (milk collection tankers and raw milk storage tanks at both plants) and at various stages of processing from raw milk to pasteurized milk (dairy I) or milk powder (dairy II). The incidence of B. cereus in the different samples was investigated, and carbohydrate metabolism and the ability to grow at 7 degrees C were assessed for the strains isolated. In total, 507 presumptive B. cereus strains were isolated. The levels present in the various samples were low; the organism could only be isolated after a pre-incubation step of 6 h at 30 degrees C. According to the ISO confirmation tests and/or carbohydrate utilization patterns (API 50 CHB), 443 (87%) of the isolates were confirmed to be B. cereus. The carbohydrate patterns revealed more than 20 different B. cereus types in raw milk and the various points in the pasteurized milk production process. Strikingly, only 10 different types were identified in the milk powder processing plant. These results indicate that selection or adaptation of strains takes place in the milk production chain, because 27 different carbohydrate utilization profiles were observed for the raw milk isolates at dairy I and II and in a previous study more than 30 different types were identified on farms. No distinct relation between the biochemical characteristics and the contamination source was observed. However, it was shown that some biotypes were found in the raw milk, during processing and in the endproducts whereas other biotypes were only detected after the pasteurization process. This indicates that post-pasteurization contamination with B. cereus occurs. The presence of B. cereus was demonstrated in 35% of the raw milk samples. When the contamination level of the raw milk was low, due to seasonal effects, B. cereus could still be detected at the various processing sites and in the final products. The percentage of positive samples increased during processing up to 71% for the pasteurized milk. The levels were generally low; the MPN counts in freshly pasteurized milk ranged from 0.9 to more than 110 per 100 ml. In dairy II the percentage of positive samples increased up to 60% for the milk powder. This is partly due to concentration effects. However, these results, in addition to the outcome of the carbohydrate utilization profiles, also indicate that B. cereus can be introduced into the milk via sources other than raw milk. In this respect, the equipment is probably an important source of contaminatisn. Only a few strains (6%) isolated at dairy I were able to grow at 7 degrees C. Since psychrotrophs have a significant impact on the keeping quality of pasteurized milk(products) the findings in this study suggest that adaptation or more likely selection of strains to growth at low temperatures occurs during storage and distribution. In the milk powder processing plant, no psychrotrophs were observed; this is due to the type of process.