Predictive microbial growth modelling for an effective shelf-life extension strategy of Chhana (Indian cottage cheese)

This study aims towards understanding microbial growth dynamics in Chhana (Indian cottage cheese) using three different primary models (viz., Logistic, Gompertz, and Barayni) for development of secondary models and relating it with the intrinsic chemical compositional properties of the samples under different storage conditions. For this purpose, Chhana was prepared in a commercial setup and shelf-life studies were performed in isothermal conditions 8 degrees C (with/without im-mersion in whey), 20 degrees C, 37 degrees C and dynamic refrigerated (8 degrees C) condition with immersion in whey. Immersion in whey was effective for extending the shelf life of Chhana. Barayni model was found better than the other models (0.94 < R2 > 0.99). Despite high correlation values all three models presented their own limitations which was then successfully overcome by novel application of three phased linearized growth model. Differential secondary models were successfully developed, and goodness of fit was evaluated with RMSE (0.032-0.084), and prediction rates achieved as high as 96.2%. The developed methods of data analysis and methods of modelling wild microbial population presented in this research would be beneficial in developing valuable insights for estimation/prediction of product shelf-life in a real-life industrial situation.

Automated summary

This study aimed to understand the microbial growth dynamics in Chhana (Indian cottage cheese) by using three different primary models (Logistic, Gompertz, and Barayni) to develop secondary models and relate them to the chemical composition properties of the samples under different storage conditions. Chhana was prepared in a commercial setup and its shelf-life was tested under various temperature conditions. Immersion in whey proved effective in prolonging the shelf-life of Chhana. The Barayni model was found to be the most accurate, and the limitations of all three models were successfully overcome by a novel three-phase linearized growth model.