Microbial Profiling, Metabolic Adaptation, and Growth Kinetic Modeling of Intact and Mechanically Damaged Milk Kefir Grains During Successive Batch Fermentations

ABSTRACT

Milk kefir grains are complex, structured microbial consortia whose biomass accumulation and fermentative performance are influenced by substrate composition, cultivation conditions, and physical integrity. Understanding their responses to mechanical disruption is essential for optimizing grain propagation and maintaining the stability of successive fermentation processes. This study investigated and compared the biomass growth, kinetic behavior, and metabolic adaptation of intact and mechanically damaged milk kefir grains during 14 successive 48‐h batch fermentation cycles (672 h) in a supplemented milk substrate previously optimized for grain biomass production. Mechanical damage was introduced at 336 h to simulate physical disruption during grain handling and separation. Grain wet mass was normalized to a common initial reference of 2.0 g, and growth kinetics were evaluated using natural‐log‐transformed biomass data. Eight candidate mathematical models were evaluated for intact grains and four for damaged grains using residual‐based goodness‐of‐fit indicators and information criteria reconstructed from MicroRiskLab exports. At 672 h, intact grains reached 134.41 ± 19.65 g, corresponding to a cumulative biomass increase of 6620.6% ± 982.7%, whereas mechanically damaged grains reached 55.27 ± 4.79 g, corresponding to 2663.5% ± 239.7%. For intact grains, the Baranyi and Gompertz models exhibited the lowest corrected Akaike information criterion (AICc = −131.36), while the reduced Baranyi model provided the best fit among the evaluated candidates for damaged grains (AICc = −66.10). Progressive changes in total soluble solids, titratable acidity, and pH indicated sustained fermentative activity throughout successive propagation cycles. Amplicon sequencing additionally characterized the bacterial and fungal composition of the source kefir grains, providing complementary microbiological information. Mechanical disruption was associated with reduced subsequent biomass accumulation and altered growth trajectories, although damage severity was not quantitatively assessed. Overall, the findings highlight the importance of preserving grain structural integrity during repeated propagation and demonstrate the potential of kinetic modeling to characterize growth responses under different physical conditions. Model‐selection results were checked by the authors in MicroRiskLab; the comparisons describe in‐sample fit and do not constitute independent predictive validation.