Mostrar registro simples

dc.contributor.advisorPelissari, Allan Libanio, 1987-pt_BR
dc.contributor.authorBarcellos, André Mendes, 1992-pt_BR
dc.contributor.otherCysneiros, Vinicius Costa, 1987-pt_BR
dc.contributor.otherUniversidade Federal do Parana. Setor de Ciencias Agrárias. Curso de Graduação em Engenharia Florestalpt_BR
dc.date.accessioned2022-02-15T18:09:54Z
dc.date.available2022-02-15T18:09:54Z
dc.date.issued2020pt_BR
dc.identifier.urihttps://hdl.handle.net/1884/73409
dc.descriptionOrientador : Prof. Dr. Allan Libanio Pelissaript_BR
dc.descriptionCoorientador : Dr. Vinicius Costa Cysneirospt_BR
dc.descriptionMonografia (graduação) - Universidade Federal do Paraná, Setor de Ciências Agrárias, Curso de Graduação em Engenharia Florestalpt_BR
dc.descriptionInclui referências: p. 31-41pt_BR
dc.description.abstractResumo : O conhecimento do volume de madeira disponível em uma floresta é informação importante para nortear as ações de manejo florestal. Entretanto, diversas são as formas de obtenção desta variável. O objetivo deste trabalho foi comparar metodologias para estimativa do volume de madeira da espécie Dalbergia nigra (Vell.) Allemao ex Benth., o jacarandá-da-baía, no bioma Mata Atlântica, visando gerar conhecimento para nortear a seleção de métodos para estimativa volumétrica. A amostra contém 29 indivíduos identificados em unidades amostrais distribuídas pelo Estado do Rio de Janeiro nas fitofisionomias Floresta Ombrófila Densa e Floresta Estacional Semidecidual. Foram coletados os diâmetros a 1,3 m de altura, alturas dos fustes, alturas totais e determinados os volumes dos fustes através do método de Smalian. Para a cubagem, foram coletados os diâmetros a 0,3, 0,5, 1,0 e 2,0 m de altura, bem como a cada um metro de altura a partir deste até a base da copa, obtidos por método não-destrutivo através do uso de técnicas e equipamentos de escalada. O processamento dos dados ocorreu no software estatístico R, sendo ajustados três modelos volumétricos clássicos por meio da análise de regressão linear: Husch, Spurr e Schumacher-Hall. O melhor foi selecionado com base nos critérios estatísticos: maior coeficiente de determinação ajustado (R²aj.), menor erro padrão da estimativa em porcentagem (Syx%) e ausência de tendências na análise gráfica dos resíduos, bem como a normalidade de suas distribuições através do teste de Shapiro-Wilk e de homocedasticidade pelo teste de Breusch-Pagan, ambos ao nível de 95% de probabilidade. O modelo de Spurr obteve os melhores valores de R²aj. e de Syx% (R²aj. = 0,9356; e Syx% = 21,89%), entretanto, apresentou distribuição não-normal dos resíduos, característica que pode resultar em estimativas não confiáveis para a variável dependente. Deste modo, o modelo de Schumacher-Hall (R²aj. = 0,9320; e Syx% = 22,48%) foi selecionado para a comparação com os métodos de estimativa a partir dos fatores de forma médio e igual a 0,7, valor usual nas ciências florestais. O teste de Breusch-Pagan retornou a presença de homocedasticidade para todos os modelos. Para as estimativas através do uso dos fatores de forma igual a 0,7 e médio, foi inicialmente obtido o volume do cilindro de cada árvore, tendo como base o d e a altura do fuste, bem como o fator de forma artificial de cada árvore, visando obter o fator de forma médio por meio da média aritmética de todos os fatores de forma da amostra. O fator de forma médio de 0,65. A comparação entre as metodologias e os volumes observados foi feita através do teste estatístico não-paramétrico de Kruskal-Wallis ao nível de 5% de significância. Visto não haver diferenças estatisticamente significativas entre os volumes observados e os volumes estimados pelas metodologias, pode ser preconizado uso do fator de forma igual a 0,7 para a espécie na área de estudo. Como vantagem, não é necessária a obtenção da altura total, variável dependente no método de regressão, bem como do volume do fuste para a modelagem e obtenção do fator de forma médio.pt_BR
dc.description.abstractAbstract : The knowledge of the wood volume available in a forest is important information to guide forest management actions. However, there are several ways to obtain this variable. The objective was to compare methodologies for estimating the wood volume of the species Dalbergia nigra (Vell.) Allemao ex Benth., the Baía rosewood, in the Atlantic Forest biome, aiming to generate knowledge to guide the selection of methods for volumetric estimation. The sample contains 29 individuals identified in sample units distributed throughout the State of Rio de Janeiro in the phytophysiognomies Dense Ombrophilous Forest and Seasonal Semideciduous Forest. The diameters at 1.3 m height, stems heights, total heights were collected and volumes of the stems were determined using Smalian's method. For cubing, diameters were collected at 0.3, 0.5, 1.0 and 2.0 m height, as well as every one-meter in height from this to the base of the canopy, obtained by a non-destructive method using climbing techniques and equipment. Data processing occurred in the R statistical software and three classic volumetric models were adjusted through linear regression analysis: Husch, Spurr and Schumacher-Hall. The best was selected based on statistical criteria: higher adjusted coefficients of determination (R²adj.), lower standard error of the estimate in percentage (Syx%) and absence of trends in the graphic analysis of the residues, as well as the normality of their distributions through the Shapiro-Wilk test and homoscedasticity using the Breusch-Pagan test, both at the 95% probability level. The Spurr model obtained the best values of R²adj. and Syx% (R²adj. = 0.9356; and Syx% = 21.89%), however, presented a non-normal distribution of residues, a characteristic that may result in unreliable estimates for the dependent variable. In this way, the Schumacher-Hall model (R²adj. = 0.9320; and Syx% = 22.48%) was selected for comparison with the estimation methods based on the average form factors and equal to 0.7, usual value in the forest sciences. The Breusch-Pagan test returned the presence of homoscedasticity for all models. For the estimates using form factor equal to 0.7 and average, the volume of the cylinder of each tree was initially obtained, based on the DBH and the height of the stem, as well as the artificial form factor of each tree, aiming to obtain the average form factor by the arithmetic mean of all form factors in the sample. The average form factor was 0.65. The comparison between the methodologies and the observed volumes was made using the Kruskal-Wallis non-parametric statistical test at 5% of significance level. Since there are no statistically significant differences between the observed volumes and the volumes estimated by the methodologies, the use of form factor equal to 0.7 for the species in the study area can be recommended. As an advantage, it is not necessary to obtain the total height, a dependent variable in the regression method, as well as the volume of the stems for modeling and obtaining the average form factor.pt_BR
dc.format.extent1 arquivo (46 p.) : PDF.pt_BR
dc.format.mimetypeapplication/pdfpt_BR
dc.languagePortuguêspt_BR
dc.subjectJacarandápt_BR
dc.subjectJacarandá - Mediçãopt_BR
dc.subjectArvores - Mediçãopt_BR
dc.subjectÁrvores - Medição - Modelos matemáticospt_BR
dc.titleAvaliação de modelos volumétricos e fator de forma para estimativa do volume comercial de Jacarandá-da-Baía (Dalbergia nigra (Vell.) Allemao ex Benth) em florestas nativaspt_BR
dc.typeMonografia Graduação Digitalpt_BR


Arquivos deste item

Thumbnail

Este item aparece na(s) seguinte(s) coleção(s)

Mostrar registro simples