Linear mixed model applied to evaluate the basic wood density of two species and a hybrid of Eucalyptus

• Autor(a) principal: De Oliveira, Rogério Antonio [UNESP]
• Data de Publicação: 2021
• Outros Autores: Batistela, Gislaine Cristina [UNESP], Simões, Danilo [UNESP], Padovani, Carlos Roberto [UNESP]
• Tipo de documento: Artigo
• Idioma: por
• Título da fonte: Repositório Institucional da UNESP
• Texto Completo: http://dx.doi.org/10.18671/SCIFOR.V49N129.03; http://hdl.handle.net/11449/208625
• topic: Analysis of regression; Repeated measures; Wood quality

The quality of products derived from reforestation woods is directly linked to the density observed in tree trunks. Due to the natural correlation between the densities collected along the longitudinal trunk discs, linear mixed models were adjusted to better study the vertical dependence of the wood basic density on three eucalyptus: Eucalyptus grandis, Eucaluptus saligna and a hybrid version of Eucalyptus. grandis x Eucalyptus urophylla. The species Eucalyptus. grandis has shown the greatest variability in the density of the discs along the trunk and also the lowest densities in relation to the other species. The hybrid had the highest density in the discs, although there was not much variability along the trunk. The linear mixed effects model proved to be one of the statistical tools for data analysis with repeated measurements, since the correlation between measurements of basic densities along the trunk should always be considered in the statistical analysis of the data. The unstructured correlation matrix was the best suited to the heterogeneity of wood density observed in the discs along the tree trunk. The interpretation of the model showed that if the intention is to produce reforestation trees with higher basic wood density and lower variability along its trunk, the hybrid is the best choice when compared to the other two species.