| Nome: | Descrição: | Tamanho: | Formato: | |
|---|---|---|---|---|
| 7.18 MB | Adobe PDF |
Orientador(es)
Resumo(s)
The geographical origin of stingless bee honey influences its physicochemical, biochemical, and sensory characteristics. In this study, we propose a chemometric approach for the geographical classification and authentication of Brazilian stingless bee honey based on melissopalynological profiling. A dataset of honey samples from five Brazilian states (Par´a, Maranh˜ao and Amazonas in the North; Rio Grande do Sul and Paran´a in the South) and five genera of stingless bees was assembled. Pollen types were identified and quantified under photonic microscopy, and the relative frequencies of key pollen types were used as input variables. Principal Component Analysis (PCA) revealed clear geographical differentiation, with Northern honeys associated with pollen types such as Tapirira guianensis, Triplaris, and Protium, whereas Southern were differentiated from Eucalyptus, Allophylus edulis, and Mimosa scabrella. Partial Least Squares Discriminant Analysis (PLS-DA) modeling achieved a satisfactory classification between Northern and Southern honeys, with zero misclassifications. In addition, a one-class classification using Data-Driven Soft Independent Modeling of Class Analogy (DD-SIMCA) successfully authenticated honeys from Rio Grande do Sul, yielding perfect metrics in sensitivity, specificity, and accuracy. These findings indicate that pollen composition, coupled with robust chemometric analyses, provides reliable
markers for the geographical classification and authentication of stingless bee honey.
Descrição
Palavras-chave
Geographic origin Multivariate analysis SIMCA PLS-DA Brazil
Contexto Educativo
Citação
HERNÁNDEZ ZARTA, Héctor Hernán [et al.] (2026) - From pollen to patterns: authentication and geographical classification of stingless bee honey by melissopalynology and chemometrics. Food and Humanity. 7:101525. DOI: 10.1016/j.foohum.2026.101525
Editora
Elsevier
Licença CC
Sem licença CC