Crescimento de espécies florestais madeiráveis como subsídio para o manejo florestal na Amazônia ocidental

• Ano de defesa: 2013
• Autor(a) principal: Cunha, Thiago Augusto da
• Tipo de documento: Tese
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Santa Maria; BR; Recursos Florestais e Engenharia Florestal; UFSM; Programa de Pós-Graduação em Engenharia Florestal
• Palavras-chave em Português: Incremento de árvores individuais; Modelo linear generalizado; Eficiência do crescimento; Autoecologia; Individual tree-increment; Generalized linear model; Gowth efficiency; Autoecology; CNPQ::CIENCIAS AGRARIAS::RECURSOS FLORESTAIS E ENGENHARIA FLORESTAL
• Link de acesso: http://repositorio.ufsm.br/handle/1/3752

Reliable growth data from trees are very important in the context of forest management. Tree-characteristics such as architecture and competition are associated with diameter growth. Previous studies analyzing tree growth changes have indicated that these variables can describe these effects more effectively. However, the precise role of these effects in the growth modeling of tropical trees needs to be studied further to discuss more effectively the actual forest management practices. Here, we reconstructed the periodic annual basal area increment (IPAg) of four long-lived timber species from the Amazon forest: Cedro (Cedrela odorata); Cerejeira (Amburana cearensis); Copaíba (Copaifera paupera) e Mogno (Swietenia macrophylla) aiming to explain the IPAg variation by developing a individua-tree growth model using the following potential predictors: tree size, morphometric data, competition; crown position and liana load. The model was fitted using a Gamma distribution in the context of Generalized Linear Models. The statistics criteria and residual distributions showed that growth model was adequate to adjust the observed IPAg data and revealed some dendrometric variables that were involved into the effects that lead the observed variation in IPAg, presenting a broad range of the observed data in situations were growth is complex and variable between species. Trees sampled in good site conditions - expressed as a good crown exposure to light, and therefore low competition - presented higher growth rates than trees sampled under high competition from neighboring trees. Besides, the variables Liana load, for Cedrela, the crown diameter and the Hegyi competition index was demonstrated to be highly influential predictive variable. Also, trees sampled in densities greater than 25 m2.ha-1 had their IPAg% significantly reduced. In the context of basal area increment yield by the space occupied in the forest (crown efficiency) the species presented mostly efficiency when growth in dominant position with thin and long crown (crown formal). Among the species, Cedrela odorata was the most efficient in growth. This behavior reinforces the importance of release trees to obtain better crown exposure.