Screening Pinus taeda (loblolly pine) families for physical and mechanical properties using vibrational spectroscopy

Abstract

In a bid to control the loblolly pine decline complex, stakeholders are using the selection and deployment of genetically superior families that are disease tolerant. It is vital that we do not compromise other important properties while breeding for disease tolerance. In this preliminary study, near infrared spectroscopy was utilized in conjunction with data collected via conventional methods to develop partial least squares regression models that were used to rapidly predict the basic density, stiffness and ultimate strength of 14 genetically superior loblolly pine families that have been selectively bred to be disease tolerant. Calibration and independent validation data were from southern pines acquired from a commercial sawmill. Seven or eight latent variables were used in model development. The coefficients of determination of the predictive models were more promising for MOE (0.58) and MOR (0.42).