Assessment of the early signs of decay of Populus deltoides wafers exposed to Trametes versicolor by near infrared spectroscopy

The use of near infrared (NIR) spectroscopy to predict levels of white-rot degradation in Populus deltoides (cottonwood) by Trametes versicolor over the couse of 1–8 days was investigated. NIR spectra were collected from cross-sectional faces following laboratory soil block decay tests. Spectra in t...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Holzforschung 2012-05, Vol.66 (4), p.515-520
Hauptverfasser: Green, Benny, Jones, Paul David, Nicholas, Darrel D., Schimleck, Laurence R., Shmulsky, Rubin, Dahlen, Joseph
Format: Artikel
Sprache:eng
Schlagworte:
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:The use of near infrared (NIR) spectroscopy to predict levels of white-rot degradation in Populus deltoides (cottonwood) by Trametes versicolor over the couse of 1–8 days was investigated. NIR spectra were collected from cross-sectional faces following laboratory soil block decay tests. Spectra in the range of 1100–2400 nm were calibrated in terms of mass loss, compression strength and exposure period based on data of standard methods. The first and second derivatives of NIR spectra were also included into the calibration, which was done by partial least squares regression. The best calibrations with the strongest R2 values were obtained in terms of exposure period (R2 0.92, second derivative), mass loss (R2 0.89, first derivative) and compression strength (R2 0.51, second derivative). As far as the validation (prediction) statistics are concerned, the following data were obtained: exposure period (R2=0.71; RPDp 1.81, second derivative), mass loss prediction (R2=0.43; RPDp=0.67, first derivative) and compression strength R2=0.28; RPDp=1.14 (second derivative). The weak statistical data can be interpreted only in a way that the data are not robust and thus an early prediction of fungal attack by NIR spectroscopy is not yet reliable.
ISSN:0018-3830
1437-434X
DOI:10.1515/hf-2012-0501