A Multiplatform Chemical Perspective on Coffee Identity: Interplay Between Origin, Cultivar, and Fermentation Time

Abstract Green coffee bean chemistry is shaped by the geographic origin, plant genetics, and post-harvest processing, but their relative contributions across complementary chemical layers, within a single, tightly controlled context, remain comparatively underexplored. We applied a multiplatform strategy utilizing ICP-based multielement profiling, UHPLC-HRMS/MS polyphenolic analysis, 1 H NMR metabolomics to investigate the contributions of farm, cultivar, and fermentation time of green coffee. Eighteen Coffea arabica microlots from three farms in Nueva Segovia, Nicaragua, were analyzed, comprising three cultivars (Maragogype, Catuai Rojo, and Bourbon) and two fermentation times (12 h and 24 h), in a balanced factorial design disentangling farm and cultivar effects. Farm of origin emerged as the dominant, significant source of variation, cultivar showed only a marginal trend and fermentation produced no significant multivariate shift. A sPLS-DA model achieved robust farm classification under leave-one-condition-out cross-validation, even among geographically close sites, whereas cultivar discrimination was weaker. Although fermentation did not shift the overall multivariate profile, mixed-effects models controlling for farm and cultivar revealed a reproducible and significant effect on a subset of features, including polyphenolic remodelling and, notably, changes in phosphorus, boron, and barium, to our knowledge, the first evidence that fermentation can influence coffee’s elemental profile. These findings indicate that different chemical layers encode environmental, genetic, and processing information with distinct sensitivities. Given the restriction to a single micro-region and harvest season, and substantial unexplained residual variance, these results are a proof-of-concept framework rather than a generalizable model; broader validation is needed before routine traceability applications.