Diagnose nutricional de minicepas e previsão de enraizamento de miniestacas de eucalipto por redes neurais

• Ano de defesa: 2010
• Autor(a) principal: Cunha, Ana Catarina Monteiro Carvalho Mori da
• Orientador(a): Paiva, Haroldo Nogueira de
• Banca de defesa: Santana, Reynaldo Campos , Gomes, José Mauro
• Tipo de documento: Tese
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Viçosa
• Programa de Pós-Graduação: Doutorado em Ciência Florestal
• Departamento: Manejo Florestal; Meio Ambiente e Conservação da Natureza; Silvicultura; Tecnologia e Utilização de
• País: BR
• Palavras-chave em Português: Nutrição florestal; Minijardim clonal; Propagação de plantas
• Palavras-chave em Inglês: Forest nutrition; Minihedges clonal; Plant propagation
• Área do conhecimento CNPq: CNPQ::CIENCIAS AGRARIAS::RECURSOS FLORESTAIS E ENGENHARIA FLORESTAL::SILVICULTURA
• Link de acesso: http://locus.ufv.br/handle/123456789/550

Eucalypt cultivation has a great economic, environmental and social importance. In order to improve its nutritional management, nutritional diagnosis can be used based on tissue analysis. Nutritional data are also useful in research which aimed at improving the management of the nursery, leading to improve productivity and quality of seedlings, as well as minimize the use of inputs. This study aimed to perform a nutritional diagnosis status of ministumps eucalypt by Kenworthy Balanced Index (IKW) and Integrated System of Diagnosis and Recommendation (DRIS) methods and to optimize and to evaluate the efficiency of artificial neural networks in order to perform nutritional diagnosis status of ministumps and to predict the rooting of minicuttings of eucalypt, based on information from nutritional monitoring status of eucalypt clonal nursery and climatic variables. It was concluded that: a) standards were obtained for methods that assess the balance sheet (Kenworthy) and balance (DRIS), b) based on the results general rules can be used to obtain reference values at diagnosis of nutritional status of eucalypt ministumps and estimate ranges of sufficiency of leaf contents, c) the universality degree of the standards varies with the methods Kenworthy, DRIS and Potential of Response to Fertilization and with the procedure used in its evaluation, d) the frequency of concordant diagnosis was lower for Eucalyptus urophylla clones, compared with Eucalyptus grandis x Eucalyptus urophylla clones, e) in terms of balance and equlibrium, the most limiting nutrients for the lack for eucalypt ministumps were Ca, Mg, Zn, Cu and B and more restrictive for excess were S, Fe and Mn, f) the results showed that artificial neural networks Perceptrons multilayer (MLP) are able of performing the task of pattern classification (diagnosis of nutritional status ministumps) and serve to obtain approximation of functions in order to predict rooting of cuttings of eucalypt.