Development of near infrared reflectance analysis calibrations for estimating genetic parameters for cellulose content in Eucalyptus globulus

Determining kraft pulp yield in the traditional way is slow and expensive, limiting the numbers of samples that may be processed. An alternative is to use a secondary standard, such as cellulose content of the wood, which is strongly correlated with kraft pulp yield. The feasibility and efficiency of predicting cellulose content using near infrared reflectance (NIR) analysis was examined for Eucalyptus globulus Labill. Calibrations for NIR prediction of cellulose content indicated that NIR analysis could be used as a reliable predictor. Standard errors of calibration were 1% or lower, and there was excellent agreement between laboratory and predicted cellulose values. Cellulose content was under moderate genetic control (h(2) ranging from 0.32 to 0.57), and genetic correlations with tree diameter and basic density were variable (ranging from -0.11 to -0.51 and -0.33 to 0.67, respectively). The advantages, disadvantages, and potential applications of NIR analysis for predicting cellulose content are examined.