Predicting microbial growth: the consequences of quantity of data.

Having developed a mathematical model of the growth responses in a laboratory medium of salmonellae (mixed inoculum of Salmonella thompson, S stanley and S. infantis) as affected by pH level, NaCl concentration and storage temperature, the consequences of systematic removal of data has been examined. Three-dimensional plots of fitted response surfaces for the whole data set and the data reduced by three strategies highlighted differences between models and illustrated the consequences of using insufficient data. The risk of an erroneous model was demonstrated.