Abstract
Oolong tea flavor is shaped by metabolite variations induced by season, processing, and cultivar. This study investigated the metabolic basis of sensory characteristics in Yongchun Foshou (YCFS) tea by integrating HS-SPME/GC-MS, HPLC, quantitative descriptive analysis (QDA), and chemometrics. Orthogonal partial least squares discriminant analysis (OPLS-DA) pinpointed key differential compounds between spring (YCFS-S) and autumn (YCFS-A) teas. Two-way ANOVA with effect size analysis confirmed that season was the dominant factor influencing most metabolites, although grade was the primary driver for free amino acids. YCFS-S contained significantly higher levels of taste-related phytochemical compounds, whereas YCFS-A exhibited elevated tea polyphenols. Statistical sensory analysis further identified a subset of key aroma compounds as drivers of seasonal and grade-based aroma differences. Correlation network analysis pinpointed 27 key sensory drivers: 12 aroma compounds defining a complex aroma profile dominated by floral and herbal notes, and 15 phytochemical compounds responsible for the umami, overall richness, and balanced bitterness. These findings were synthesized to construct the first dedicated flavor wheel for YCFS, elucidating how seasonality and grade collectively shape its flavor profile and providing a scientific basis for quality control and standardized production.