Técnicas para predição de variáveis dendrométricas em sistema de integração lavoura-pecuária-floresta

• Ano de defesa: 2018
• Autor(a) principal: Cerqueira, Clebson Lima
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Mato Grosso; Brasil; Faculdade de Engenharia Florestal (FENF); UFMT CUC - Cuiabá; Programa de Pós-Graduação em Ciências Florestais e Ambientais
• Palavras-chave em Português: Biometria florestal; Modelagem de efeitos mistos; Eucalipto; CNPQ::ENGENHARIAS; Forest biometry; Mixed effects models; Eucalyptus
• Link de acesso: http://ri.ufmt.br/handle/1/2945

The integration Crop, Livestock and Forest systems is a strategy to intensify use of soil with a difersification of profit to producer with enviromental earnings. However, the studies to evaluate the forest production in this system are scarce, arising the necessity and interesting by the forest producer knowing the particularities of forest's grow and forest's development in this systems. From now on, it is necessary the development of studies with aplication of convencional modeling and the technique of mix effects for eucalyptus in ILPF where are used diferentes configuration of spacing a space arrangement. The objective of this study, was to evaluate estimate obtained at the taper models not-segmented, segmented and variable form; verifying the effect of spacing and arrangement of stand at trunk's form, volume, and trees' heigt; Evaluating the potencial of mix effects modeling to predicted dendrometrics variables. The data was obtained at a integrated system, in an experimental area implanted by company empresa Brasileira de Pesquisa Agropecuária (EMBRAPA) Crop, Livestock and Forest, loacalized at Sinop, in the state of Mato Grosso. In this study, was used four treatments: Presence of peripheral trees at pasture (T1), area of pasture fully wooded (T2), Livestock and Forest systems (T3) and homogeneous planting (T4). The data was obtained by Log volume by direct measurement in felled trees in each treatment, in a total of 186 trees with 51 months of age. To height modeling, volume and taper was used two techniques of modeling, that was tradicional regression and mix effects modeling. To evaluate if there is influence of the treatments at the fit of hypsometric, volumetric and taper equations was adopted the identity test. With base in the evaluating statistics criteria of fit was verify that the variable taper models are more accurated to predict the taper and volume and the Perez's model was the most suitable. The spacing and arragement of stand influence significantly at the height, volume and taper. The mix effects modeling presents a potencial to predict dendrometrics variables, as well as to calibrate models with grouped data of different forest conditions.