To elucidate the molecular basis of the functional deterioration of Qula-derived casein, proteomic analysis was conducted, and the protein composition, structural characteristics, and functional properties of imported casein, fresh yak milk casein, and Qula-derived casein were systematically compared. Structural characterization was performed by spectroscopic and microscopic analyses, and physicochemical properties were measured in parallel. Qula-derived casein contained higher proportions of αs1-casein than imported casein and β-casein than both imported and fresh yak milk casein, but a lower proportion of κ-casein than either sample. Relative to fresh yak milk casein, Qula-derived casein showed decreases of 31.8% and 12.3% in α-helix and β-sheet contents, respectively, accompanied by a 22.3% increase in random coil content. Its microstructure also changed from a loose, porous morphology to dense, block-like aggregates. Accordingly, its emulsifying activity index, emulsion stability index, solubility, and surface hydrophobicity decreased to 29.3 m²/g, 23 min, 0.6%, and 199, respectively, representing reductions of 36.7%, 25.8%, 33.3%, and 32.1% compared with fresh yak milk casein. These findings indicate that alterations in casein isoform composition are associated with protein unfolding, aggregation, and the deterioration of functional properties during Qula processing. The new knowledge of this study lies in providing insight into the molecular mechanism by which traditional Qula processing may induce casein subtype changes, unfolding, and aggregation, and in describing a possible association between decreased κ-casein abundance and impaired functionality. These findings support the value-added utilization of Qula-derived casein and the optimization of traditional processing.

To study the relationship between the microbial community structure and the quality of alfalfa silage by adding alfalfa-fermented green juice (aFGJ), the Box–Behnken design was used to optimize the process of preparing aFGJ. High-throughput sequence technology was used to detected the microbial community structure of aFGJ and silage samples, and two silage experiments were designed: silage with aFGJ (LZQC) and silage with no addition (CK). In addition, the chemical composition and fermentation quality of silage samples at different times were determined. The result showed that the optimal aFGJ culture condition for LAB quantity was an addition of 2.6% glucose to alfalfa green juice and an anaerobic culture at 35 °C for 49 h, and the number of viable bacteria can reach 8.0 × 10⁸ cfu·mL⁻¹. The main microorganisms (species level) were Lactobacillus brevis (57.63%) and Lactobacillus plantarum (37.31%). There was a significant increase in the relative abundance of Pediococcus. pentosaceus and L. plantarum and a reduction in L. brevis after silage for 70 days with the added aFGJ; meanwhile, the chemical composition (DM losses, aNDF) and fermentation quality (pH, NH₃-N/TN, LA) were significantly better than that of the CK group. The addition of aFGJ changed the types and abundance of microbial communities, increased the number of beneficial lactic acid bacteria (LAB), and decreased the number of harmful bacteria, which had a significant effect on the quality of the silage.