Multisite evaluation of the 3-PG model for the highest phenotypic plasticity Eucalyptus clone in Brazil

Different ecophysiological models are used to predict the productivity of forest plantations worldwide. The Physiological Principles in Prediction Growth (3-PG) model has been successfully used for this purpose since 1997. In this study, the 3-PG model was parameterized and validated to predict the productivity of the most planted clonal eucalypts in Brazil (Eucalyptus urophylla) in different regions of the country and assess the attainable productivity of this same clone for a region little exploited for this plantations in the country. Through data collection carried out between 2012 and 2018, conducted in 36 sites distributed across an environmental gradient that spans over 3500 km in Brazil, it was possible to parameterize the 3-PG model using data that represent a portion of the soil and climate diversity of forest plantations in Brazil. We determined the initial biomass of the compartments (stem, leaf, and root), relationship between net and gross productivity, average wood density, maximum stem biomass for 1000 trees ha⁻¹, maximum stomatal conductance, bark and branch fractions for the initial and mature age, specific leaf area for the initial and mature age, and allometric parameters at different ages. Considering the model initialization period, the initial age was defined at 12 months after planting and the final age was considered at 80 months. Model calibration was performed in four experimental sites (special sites for the calibration set), which correspond to the environmental diversities of the project, and model validation was performed by applying the setup of the model obtained in 10 other sites (regular sites for validation set) of the same experimental network. In all sites tested, estimates of basal area, diameter at breast height, leaf area index, and stem biomass agreed with the measured values. On average, the estimates of diameter and breast height were 4.91% higher than the observed measures, whereas the stem biomass estimates were 16.44% lower and the leaf area index was 26.22% lower; moreover, the estimates were overestimated for the first three years and underestimated in recent years. Overall, the model was able to capture the soil and climate differences for predicting E. urophylla clone productivity. Its application in a region not yet exploited for eucalypts plantations may help investors in the region select areas for acquisition, planting extensions, and extend production technologies.