Modelagem preditiva do crescimento/morte de Saccharomyces cerevisiae em co-cultura com Lactobacillus fermentum em mosto de caldo de cana-de-açucar / Predictive modeling growth/death of Saccharomyces cerevisiae in coculture with Lactobacillus fermentum in sugar cane must

AUTOR(ES)
DATA DE PUBLICAÇÃO

2008

RESUMO

In Brazil and other countries around the World, the sugar cane must fermentation for alcohol production is carried out without sterilization of the sugar cane juice, being an excellent medium for the multiplication of undesirable contaminants. Among the contaminants of sugar cane must, Lactobacillus fermentum can be considered one of the most relevant being able to cause the flocculation of yeasts and to reduce ethanol yield and productivity of the fermentation. This research aimed to predict the growth parameters lag time (l), specific growth rate (m) and maximum population (Rg) of Saccharomyces cerevisiae and Lactobacillus fermentum in individual and cocultures. Predictive modeling of microorganisms growth in co-culture and pure culture was done through the application of primary growth models of Baranyi and modified Gompertz, varying temperature (24 ? 32ºC) and inoculum concentration of L. fermentum (101 ? 108 CFU/mL) with a fixed inoculums level of S. cerevisiae (106 CFU/mL). For the primary modeling it was also, adjusted the quase-chemical model for the construction of this model Matlab software version 7.5 was used. The inoculation of the cultures was carried out in sugar cane must industrially clarified and adjusted to 21.5ºBrix, heat treated must (121ºC per 40 minutes) were inoculated with the yeast and lactobacilli, the adjustment of the inoculum was done respectively by counting the suspension in Neubauer chamber and Densimat (BioMérieux, S.a., France). For each assay carried out with the mixed culture, two others tests with pure culture were done for comparison of growth parameters. The fermentation assays were conducted in an incubator with continuous agitation (120rpm) (New Brunswick Scientific, Model G 27, U.S.A.) and controlled temperature. From the data obtained, the growth parameters were determined through the adjustment of the data with DMFIT software for Baranyi and Roberts and Modified Gompertz. No siginificant diference (p<0.10) was found for l,m and Rg for the yeast in pure culture neither for l and m for Lactobacillus fermentum, but for lactobacilli Rg was significantly different for pure a coculture. When the yeast reached the stationary phase the lactobacilli growth rate was increased and its stationary phase it was observed after ward. Baranyi and Roberts?s model best described lag phase was expected. The quasichemical model even though it describes growth/decline it does not model the stationary plateau. The secondary model for the maximum population of the lactobacilli as function of fermentation temperature and inocullum level showed that this last variable was significant and its interaction with temperature (p<0.10). For the pure culture temperature was significant on the viability index. But for mixed culture for a fixed fermentation time (21h) the lactobacilli level of contamination was highly significant. Best IV was observed for L=103UFC/mL and T=25oC. When the maximum ethanol production was modeled for T, L independent variables. The temperature was significant a maximum was observed at T=28oC, L=105UFC/mL. These two models are practical application for the alcoholic fermentation industry

ASSUNTO(S)

coculture predictive microbiology saccharomyces cerevisiae fermentação alcoolica alcohol fermentum co-cultura microbiologia preditiva lactobacillus fermentum

Documentos Relacionados