Hierarchical Chemotaxonomic Differentiation in Cannabis Chemovars Using Quantitative HPLC Cannabinoid Profiling and Multivariate Chemometrics

-tetrahydrocannabinol (THC) to cannabidiol (CBD) ratio, yielding canonical chemotypes I, II, and III. However, this binary framework overlooks the chemical diversity contributed by the minor cannabinoids. High-performance liquid chromatography (HPLC) following the AOAC Official Method 2018.10 was employed to quantify nine cannabinoids (THCA, THC, CBDA, CBD, CBGA, CBG, CBC, CBDV, and CBN) across 36 commercially and medicinally relevant cannabis varieties. Quantitative profiling revealed substantial phytochemical heterogeneity, with total THC ranging from 0.41% to 15.64% and total CBD ranging from 0.09% to 12.32% (w/w). Unsupervised principal component analysis (PCA) demonstrated that the first two principal components explained 62.7% of the total variance. PC1 (37.6%) captured the THCA-CBDA polarity axis, while PC2 (25.1%) was dominated by minor cannabinoids (CBC; loading 0.417), CBGA (0.314), and CBG (0.258). Supervised partial least squares discriminant analysis (PLS-DA) using only the nine cannabinoid variables achieved 94.2% cross-validated accuracy and 100% test-set accuracy in predicting the chemotype class, with CBC identified as the third most discriminatory variable (variable importance in projection, VIP = 1.34). Hierarchical clustering resolved three principal clades and further subdivided THC-dominant accessions into CBC-enriched (Sour Diesel, Cinderella Jack) and CBGA-enriched (Mother Gorilla, Auto Lemon Kix) subclusters. A multivariate "metabolic coordinate" system based on PC1/PC2 scores is proposed as a quantitative and reproducible alternative to the traditional Type I/II/III and sativa/indica nomenclatures. This study introduces an empirically grounded framework for variety authentication, quality control, and enhanced precision breeding in the rapidly growing medicinal cannabis sector, for both human and veterinary applications.