No abstract

Burman design was used to determine the most influential of 9 variables (glucose, lactulose, tween80, K 2 HPO 4 , sodium acetate, MgSO 4 , pH, shaking and inoculum size) on the growth of L. plantarum BH14. Glucose, lactulose, MgSO 4 and shaking had a significant effect on growth of this strain at a 70% confidence level. Glucose, lactulose and MgSO 4 showed positive coefficient but shaking had a negative coefficient, thus shaking was avoided in subsequent experiments. Composite design was used to determine the optimum concentrations of the significant variables. An optimized formulation of nutrition levels was suggested from the software at the following concentrations: 11.59 g LG 1 glucose, 11.59 g LG 1 lactulose and 0.23 g LG 1 MgSO 4. Growth of L plantarum was compared to that recorded on the reference medium 'vegetal MRS'. The results showed that viable counts in the optimized culture medium were significantly higher than those in the vegetal MRS medium.