Differentiation of leafy spurge ( Euphorbia esula L.) accessions based on chemical constitution using Curie-point pyrolysis-mass spectrometry-pattern recognition

Abstract

Curie-point pyrolysis-mass spectrometry, in combination with multivariate pattern recognition methodology (Py-MS-PR), is shown to be an effective method for differentiating accessions of leafy spurge ( Euphorbia esula L.) based on the chemical constitution of plant latex. Three standard multivariate statistical programs — hierarchical cluster analysis, factor analysis, and principal component analysis — correctly classified pyrolysis-mass spectra as either E. esula or E. cyparissias. In addition, different accessions of E. esula corresponding to three different populations (Mandan, ND; Cambridge, ID; and Baker, OR), were classified by this method. The features ( m/z) responsible for each differentiation/classification were identified using Fuzzy c-Varieties Pattern Recognition (FCV). The results of this study were compared to results obtained in an earlier study using Curie-point pyrolysis-gas chromatography-pattern recognition (Py-GC-PR) to differentiate/classify the same set of leafy spurge plants. Results from both techniques (Py-MS-PR and Py-GC-PR) elucidated various biotypes of leafy spurge, demonstrating the applicability of both methods for differentiating/ classifying this troublesome weed.