Geographic determination of Pinus ponderosa using DART TOFMS, ICP-MS, and LIBS handheld analyzer

Abstract

Due to legal requirements on international imports, it is important for law enforcement and regulatory agencies to identify the geographical provenance of timber. Current methods for geographic identification utilize data generated by direct analysis in real time time-of-flight mass spectrometry (DART TOFMS), genetics, and isotope-ratio mass spectrometry (IRMS), but identification methods based on genetics and IRMS data require months to years to create usable databases. This study used machine learning algorithms to compare the results of DART TOFMS, inductively coupled plasma mass spectrometry (ICP-MS), and a handheld laser-induced breakdown spectroscopy (LIBS) analyzer for use in geographic identification of five populations of Pinus ponderosa spaced between 14 to 72 km apart. The results of the study showed comparable performances from machine learning algorithms applied to the ICP-MS and LIBS data with accuracy and kappa values over 90% while the DART TOFMS had an accuracy of 76% and a kappa value of 70%. This study demonstrated that data from the LIBS handheld analyzer is a viable and intriguing alternative to ICP-MS and DART TOFMS analyses in generating training databases and further indicates that trace elemental analysis via ICP-MS is a promising method for generating databases used to identify the origin of timber.