guest ::
| Home > Publications database > Advanced mathematical modelling of hyperspectral data to quantify the optically inactive secondary metabolites eugenol and methyl eugenol in the medicinal plant holy basil |
| Home > Publications database > Advanced mathematical modelling of hyperspectral data to quantify the optically inactive secondary metabolites eugenol and methyl eugenol in the medicinal plant holy basil |
| Conference Presentation (After Call) | FZJ-2025-04851 |
; ; ; ; ; ; ; ;
2025
Abstract: The quantification of secondary metabolite content is crucial for the industrial production of high-quality essential oil from medicinal plants, i.e. to determine an optimal harvest time to achieve the desired quality. Spectral reflectance measurements provide a potential technique to perform non-destructive and real-time monitoring of secondary metabolite content. However, not all secondary metabolites are optically active in the range of hyperspectral sensors. In this study, a method for the non-destructive quantification of eugenol (Eu) and methyl eugenol (MeEu) in holy basil (Ocimum tenuiflorum L.) was developed by combining leaf-level hyperspectral measurements with mathematical modelling based on partial least squares regression (PLSR). To generate a wide range dataset of secondary metabolite content, three experiments investigating developmental variation, osmotic stress response and secondary metabolism induction by methyl jasmonate were conducted with two commercial cultivars. Hyperspectral point measurement (350-2500 nm) of individual leaves were combined with destructive quantification of Eu and MeEu by gas chromatography-mass spectrometry. We observed variation of Eu and MeEu concentrations between cultivars, treatments, and leaf ages. PLSR modelling based on the full wavelength spectrum resulted in quantification of Eu and MeEu with R2 of 0.62 and 0.70, respectively, with RMSEP of 1.10 and 0.98. We explain the high correlation by an indirect quantification of these compounds based on associated changes in plant secondary metabolism. This study demonstrated that hyperspectral data combined with PLSR provides a promising technique for the non-destructive quantification of foliar secondary metabolites for industrial applications even in absence of specific absorption feature within the measurement range.
|
The record appears in these collections: |