The boundary for growth of Zygosaccharomyces bailii in acidified products described by models for time to growth and probability of growth

Citation
P. Jenkins et al., The boundary for growth of Zygosaccharomyces bailii in acidified products described by models for time to growth and probability of growth, J FOOD PROT, 63(2), 2000, pp. 222-230
Citations number
17
Categorie Soggetti
Food Science/Nutrition
Journal title
JOURNAL OF FOOD PROTECTION
ISSN journal
0362028X → ACNP
Volume
63
Issue
2
Year of publication
2000
Pages
222 - 230
Database
ISI
SICI code
0362-028X(200002)63:2<222:TBFGOZ>2.0.ZU;2-N
Abstract
Models to predict days to growth and probability of growth of Zygosaccharom yces bailii in high-acid foods were developed, and the equations are presen ted here. The models were constructed from measurements of growth of Z. bai lii using automated turbidimetry over a 29-day period at Various pH, NaCl, fructose, and acetic acid levels. Statistical analyses were carried out usi ng Statistical Analysis Systems LIFEREG procedures, and the data were fitte d to log-logistic models. Model 1 predicts days to growth based on two fact ors, combined molar concentration of salt plus sugar and undissociated acet ic acid. This model allows a growth/no-growth boundary to be visualized. Th e boundary is comparable with that established by G. Tuynenburg Muys (Proce ss Biochem. 6:25-28, 1971), which still forms the basis of industry assumpt ions about the stability of acidic foods. Model 2 predicts days to growth b ased on the four independent factors of salt, sugar, acetic acid, and pH le vels and is, therefore, much more useful for product development. Validatio n data derived from challenge studies in retail products from the U.S. mark et are presented for Model 2, showing that the model gives reliable, fair-s afe predictions and is suitable for use in predicting growth responses of Z . bailii in high-acid foods. Model 3 predicts probability of growth of Z. b ailii in 29 days. This model is most useful for spoilage risk assessment. A ll three models showed good agreement between predictions and observed valu es for the underlying data.