-The acidification and reduction activities of lactic starters have been followed by continuous measurement of the pH and the Eh during the course of milk fermentation. These measurements allowed the calculation of the maximum acidification and reduction rates, and the time, pH and Eh at which these values occurred. Nine strains of Lactococcus sp., 6 strains of Streptococcus thermophilus and 5 strains of Lactobacillus helveticus were studied. In general, the maximum reduction rate of the lactococci was six-fold higher than those of the streptococci and lactobacilli. On the other hand, the streptococci and the lactobacilli acidified with a higher maximum acidifying rate than those of lactococci. Consequently, it was observed that all the cultures with lactococci reached their final Eh before the end of the lactic acid fermentation, while acidifications with the streptococci or the lactobacilli finished before the end of the reduction phase. A principal components analysis clearly differentiates the three species on the basis of their aptitudes for acidification and reduction. This new approach might be used to select adequate starters for the manufacture of fermented dairy products. terminait avant la fin de la phase de réduction. L'analyse en composante principale différencie clairement les 3 espèces sur la base de leurs aptitudes acidifiantes et réductrices. Ces nouveaux paramètres pourraient être associés à ceux classiquement utilisés pour sélectionner les levains les plus adéquats pour la fabrication de produits laitiers fermentés. Acidification / réduction / Lactococcus / Streptococcus thermophilus / Lactobacillus helveticus

No abstract

Geotrichum candidum was cultivated at the surface of solid model media containing peptone to simulate the composition of Camembert cheese. The surface growth of G. candidum induced the diffusion of substrates from the core to the rind and the diffusion of produced metabolites from the rind to the core. In the range of pH measured during G. candidum growth, constant diffusion coefficients were found for lactate and ammonium, 0.4 and 0.8 cm(2) day(-1), respectively, determined in sterile culture medium. Growth kinetics are described using the Verlhust model and both lactate consumption and ammonium production are considered as partially linked to growth. The experimental diffusion gradients of lactate and ammonium recorded during G. candidum growth have been fitted. The diffusion/reaction model was found to match with experimental data until the end of growth, except with regard to ammonium concentration gradients in the presence of lactate in the medium. Indeed, G. candidum preferentially assimilated peptone over lactate as a carbon source, resulting in an almost cessation of ammonium release before the end of growth. On peptone, it was found that the proton transfer did not account for the ammonium concentration gradients. Indeed, amino acids, being positively charged, are involved in the proton transfer at the beginning of growth. This effect can be neglected in the presence of lactate within the medium, and the sum of both lactate consumption and ammonium release gradients corresponded well to the proton transfer gradients, confirming that both components are responsible for the pH increase observed during the ripening of soft Camembert cheese.

Penicillium camembertii was cultivated on a jellified peptone-lactate based medium to simulate the composition of Camembert cheese. Diffusional limitations due to substrate consumption were not involved in the linear growth recorded during culture, while nitrogen (peptone) limitation accounted for growth cessation. Examination of gradients confirmed that medium neutralization was the consequence of lactate consumption and ammonium production. The diffusion of the lactate assimilated from the core to the rind and that of the ammonium produced from the rind to the core was described by means of a diffusion/reaction model involving a partial linking of consumption or production to growth. The model matched experimental data throughout growth.