Resumen

CASTRO, Gustavo et al. Comparison of Sigmoid Models Applied to the Growth of Lactococcus lactis subsp: Lactis. Rev. Cient. (Maracaibo) [online]. 2008, vol.18, n.5, pp.582-588. ISSN 0798-2259.

Predictive Microbiology is a useful tool to describe and predict the bacterial growing in food. Milk samples from total morning milking of two herds, buffalo (n=6) and cattle (n=6), managed under similar conditions, located in Mara Municipality, Zulia State, Venezuela, were analyzed to compare the application of different sigmoid models to the growth of Lactococcus lactis subsp. lactis strain. Sanitary quality was determined by means of pH, acidity titration and detection of inhibitors. A series of assay tubes were prepared with sterilized milk at 110 ± 2°C for 10 min. The milk in the assay tubes was inoculated with Lactococcus at 1 x 10⁴ ufc/mL. Then, the assay tubes were incubated at 36 ± 0.5°C and every two hours dilutions and culture were prepared in Petrie dishes with M-17 agar. Log₁₀ of ufc/mL was determined for 0; 2; 4; 6; 8 y 10 hours post incubation. A regressions analysis with Gompertz, Logistic, Stannard and Richards models through Marquardt algorithm was used. Similarity was found when models were adjusted. The Stannard and Richards models were equal for the data estimation and they presented the greatest difficulty to reach the convergent criterion. The three parameters Gompertz model showed significant difference compare to the four parameters models. The Gompertz and Logistic models (with four parameters) modified by Gibson are considered the best to model and predict the growing of the microorganism under study.

Palabras clave : Lactococcus lactis; predicitive microbiology; Gompertz; buffalo.