Modelo para predizer a altura de clones de Eucalyptus urograndis em tabuleiros costeiros, no estado de Alagoas

Abstract

Eucalyptus plantations constitute the foundation of Brazilian forest production, distinguished by their high productivity and broad adaptability to diverse environments. In the coastal tablelands of Alagoas, the development of accurate hypsometric models is essential to optimize forest inventories and to support appropriate plantation management. In this context, the present study aimed to develop and evaluate models to predict the height of a clonal Eucalyptus urograndis stand under three spacing regimes. The experiment was conducted at the Engineering and Agricultural Sciences Campus of UFAL, in a randomized block design, with three spacings (3×2 m, 3×3 m, and 3×4 m) and two clones (VE41 and I144), totaling 2,152 valid observations between 12 and 72 months of age. The results showed considerable variation in tree growth, with mean heights of 13.97 m and mean diameters of 11.91 cm, reflecting the influence of environmental conditions and resource competition. Spacing had a significant effect on height, indicating that wider spacings favored height growth, whereas the clone factor and the clone × spacing interaction were not significant. Pearson correlations revealed strong associations between height, diameter, and age (r = 0.898 and r = 0.945, respectively), reinforcing the importance of these variables in hypsometric modeling. Several models, with and without intercepts, were fitted and evaluated using the coefficient of determination (R2) and the coefficient of variation (CV%). Among the tested equations, the simple linear expression h = 0.5213 ⋅ d + 0.1901 ⋅ tshowed the best performance, with Rଶ = 0.9929and CV = 9.19%, in addition to randomly distributed residuals and no evidence of heteroscedasticity, confirming its robustness and suitability for height prediction. The other equations, although well-fitted, included excessive variables and showed greater variance inflation, making the linear model more efficient and practically applicable in forest inventories. It is concluded that the proposed model stands out for its accuracy, simplicity, and operational applicability, and can be used in the planning and management of Eucalyptus urograndis stands under edaphoclimatic conditions similar to those of the coastal tablelands of Alagoas. The use of this model can reduce field costs, optimize growth estimates, and improve decision-making in forest monitoring activities.