The potential of Mid-Infrared spectroscopy for prediction of wood density and vulnerability to embolism in woody angiosperms

Abstract

Xylem resistance to embolism formation determines the species-specific drought tolerance and the survival prospects of plants under extreme climatic conditions. Fourier Transform-Infrared (FTIR) spectroscopy is a cost-effective and rapid analytical tool with potential beyond its current use in plant physiology. We tested the use of FTIR spectroscopy as a method for estimating wood density (WD) and xylem resistance to embolism formation (P50) in 24 angiosperm species. Higher WD was associated with more negative P50 (r2 = 0.41). Partial least squares regression was applied to establish models of FTIR spectra and the reference data. They showed a high predictive quality for WD (r2 = 0.73), whereas the prediction of P50 was weaker (r2 = 0.49). By including WD in the model as an additional factor influencing P50, its predictive power significantly increased (r2 = 0.59). The spectral range in the model elaboration has been also narrowed (bands of lignin, cellulose, hemicellulose), but this did not influence the model descriptors, suggesting that for P50 prediction broad spectral range is more informative than narrow band regions reflecting main wood constituents. In conclusion, FTIR spectroscopy associated with WD measurements has proven to be a promising alternative to traditional methods for screening of individual- or species-specific resistance to embolism in angiosperms.

Keywords: Fourier Transform-Infrared spectroscopy; PLS regression; drought resistance; vulnerability to embolism; wood density.