A critical assessment of four most probable number procedures for routine enumeration of cheese-damaging clostridia in milk

Abstract In cheese production, knowledge of clostridial spore levels in milk is crucial to avoid economic losses due to late-blowing, a characteristic quality defect observed especially in hard cheeses. However, no international standard method is available to quantify clostridial spores in milk. Since performance data on the currently applied enumeration methods have not been systematically investigated, in this study we assessed four routine media and most probable number (MPN) methods for the enumeration of cheese-damaging clostridia in milk. Among the four evaluated methods, statistically significant differences became evident, while moderate to optimal correlations were observed. Bacterial isolates from positive reactions were identified using molecular methods. All four methods assessed showed a severe lack of selectivity due to the incorporation of non-clostridial spore-formers in the counts produced. The results illustrate major drawbacks of currently used methods and stress the necessity for improving method selectivity, sensitivity and reliability before standardisation can be envisaged.