Adaptation, calibration and evaluation of a simple agrometeorological model for wood Eucalyptus productivity estimation

Abstract

The Agroecological Zone Model-FAO (AZM) makes use of a physiological basis to estimate the potential productivity (Yp) and an empirical approach to simulate the effect of water deficit on attainable productivity (Ya). The Eucalyptus genus is the most planted in Brazil, with approximately six million hectares, commanding immense economic relevance for the country. Considering the importance of this forest species and the influence that weather conditions have on its growth, the aim of this study was to adapt, calibrate and evaluate the AZM to estimate Eucalyptus productivity for eight Brazilian clones. To accomplish this, forest inventory data from 23 trials and eight Eucalyptus clones (classified as plastic, tropical and subtropical) were obtained from different growing regions in Brazil, from 2011 to 2017. The calibration and adaptation of the model resulted in a significant improvement of its performance. The root-mean-square error was approximately 110 m3 ha−1 when not calibrated and 39 m3 ha−1 after calibration. The calibration also improved precision, with R2 going from 0.73 to 0.82, accuracy, with d index increasing from 0.70 to 0.93, and confidence, with c index going from weak (c = 0.59) to very good (c = 0.84). During the evaluation of the model with independent data, its performance was classified as great (c = 0.87). The AZM, adapted to the Eucalyptus forest, presented satisfactory performance for estimating Eucalyptus wood volume per hectare, representing a useful tool for all players in the forest sector.