Cutting-edge approaches for honey authentication: Chemical, molecular,and ai-driven strategies for botanical origin verification

Auhtors : Tushar Khare, Kareem A. Mosa , Rani a Hamdy , Attiat Elnaggar , Shifa Malik, Suad Kadeem Khan, Al i El -Keblawy , Foua d Lamghari, Ahmed M.S. Alhmoudi, Khawla M. Alyammahi.

A B S T R A C T Hone y is a widely co nsume d food co mmo dit y with si gni ficant co mme rcial value, ofte n dete rmine d by it s bota n i - ca l an d ge ograp h ica l or igin. This ha s le d to issues such as fa lsified labe lin g an d adulte ration. Whil e rapi d dete c - tion method s exist, hone y authentication techniques ar e stil l evol vin g to enhanc e accuracy . This articl e review s cu rrent trends in authentica tin g ho ney's bota n ica l or igin, whic h is linked to it s pollen co ntent an d bi ochem ica l pr ofile . Tr aditional authentication method s includ e meli ssopalyno log y an d chromatographic/spectr oscopic fi n - ge rprin ting, whic h identify unique chem ica l marker s of sp ecifi c plan t species. Additionally , pollen DN A metaba r - co din g is highlighte d as a promisin g approach , usin g plan t -specific DN A ba rcode s fo r pr ecise bota n ica l identifica - tion . Th e review also emph asize s th e impo rtanc e of chem ome tri c -multivariate anal ysi s fo r pr ocessin g larg e datasets ge nerated by thes e methods. Fu rthermore , artificial inte lligenc e -driven machin e lear nin g tool s have show n near -complete accuracy in classifyin g hone y sa mples an d detectin g adulte ration. Despit e thes e advanc e - ments, challenges pe rsist , includin g inco mplet e re ference databases, no n -standardized pr otocols , an d th e high cost of advanced methodol ogies , li mitin g acce ssibi lit y fo r smal l -scal e be ekeepers. To addres s thes e gaps , th e arti - cl e advocate s fo r co lla b orative , mu ltidi scipl inary research to refine authentication techniques , expand open - access re ference libraries, an d develo p affordable , user -friendly se nso r technologies . Such effort s ai m to enhanc e co nsume r co nfidenc e an d maintain ma rke t integrit y by ensu rin g accurate hone y authentication .

Overview

Honey’s commercial and medicinal value depends heavily on its botanical and geographical origin, making it vulnerable to fraud and mislabeling. This comprehensive review by Khare et al. (2025) explores advances in chemical, molecular, and AI-driven strategies for verifying honey authenticity. Traditional approaches such as melissopalynology are being complemented—and in some cases replaced—by spectroscopic, chromatographic, DNA-based, and artificial-intelligence methods that enhance accuracy and reproducibility.

Traditional Methods

• Melissopalynology (microscopic pollen identification) remains the reference method for determining floral origin.Limitations: time-consuming, subjective, and requires expert knowledge and robust pollen libraries.

• Chromatography (HPLC, GC-MS) and spectroscopy (FTIR, NMR, Raman, UV–Vis, NIR) offer detailed chemical “fingerprints.”

  • FTIR and Raman spectroscopy, when paired with chemometrics, reach up to 97 % classification accuracy for floral origin.

• Electrophoresis (SDS-PAGE) and elemental profiling (ICP-MS, AAS) detect protein or mineral signatures indicative of specific floral sources.

Chemical and Chemometric Approaches

Distinct chemical markers—phenolics, flavonoids, quinochalcones, terpenes, and glycosides—enable differentiation between monofloral honeys:

• Kaempferitrin → Camellia oleifera honey

• Lumichrome → Centaurea cyanus honey

• (−)-Gallocatechin gallate → Triadica cochinchinensis honey

• Hydroxysafflor yellow A → Carthamus tinctorius honey

• Agnuside → Vitex agnus-castus honeyDNA Barcoding pagination_YJFCA_…

Chemometric tools such as PCA, LDA, PLS-DA, and k-NN process high-dimensional datasets from these analyses, improving discrimination between honey types and regions.

Molecular DNA and Metabarcoding Techniques

DNA metabarcoding uses high-throughput sequencing (HTS) of pollen DNA (e.g., ITS2, rbcL, trnL, matK) to identify floral species within honey samples.

• Demonstrated 64 % repeatability vs. 28 % for traditional microscopy.

• Detects taxa even in processed or degraded honeys.

• Challenges: incomplete barcode databases (BOLD, PLANiTS, ITS2), low DNA yield, and cost barriers.Emerging environmental DNA (eDNA) approaches show promise for linking hive or landscape biodiversity to honey origin but remain underexplored.

Artificial Intelligence and Machine Learning

AI models integrated with spectroscopy or chemical datasets dramatically enhance classification and fraud detection:

• Support Vector Machines (SVM), Convolutional Neural Networks (CNN), and Neural Networks achieve >95 % accuracy in differentiating floral types and detecting adulteration.

• AI fusion of spectral data and pollen databases offers near-real-time authentication, suggesting potential for portable, field-ready sensors.

Key Challenges & Future Directions

• Standardization: Lack of unified global protocols for sampling, data interpretation, and reference libraries.

• Accessibility: High costs of advanced tools limit use by small beekeepers.

• Integration: Combining chemical, molecular, and AI-based systems could yield affordable, automated honey authentication kits.

• Collaboration: Multidisciplinary cooperation is essential to develop open-access databases and IoT-linked sensors for global traceability.

Reference: Cutting-edge approaches for honey authentication: Chemical, molecular, and AI-driven strategies for botanical origin verification. Journal of Food Composition and Analysis, 107974. https://doi.org/10.1016/j.jfca.2025.107974