Tamanho da amostra para seleção de famílias de cana-de-açúcar

• Ano de defesa: 2007
• Autor(a) principal: Leite, Mauro Sergio de Oliveira
• Orientador(a): Peternelli, Luiz Alexandre
• Banca de defesa: Silva, Derly José Henriques da , Martins Filho, Sebastião
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Viçosa
• Programa de Pós-Graduação: Mestrado em Genética e Melhoramento
• Departamento: Genética animal; Genética molecular e de microrganismos; Genética quantitativa; Genética vegetal; Me
• País: BR
• Palavras-chave em Português: Cana-de-açúcar; Seleção; Número de plantas; Simulação
• Palavras-chave em Inglês: Sugarcane; Selection; Number of plants; Simulation
• Área do conhecimento CNPq: CNPQ::CIENCIAS AGRARIAS::AGRONOMIA::FITOTECNIA::MELHORAMENTO VEGETAL
• Link de acesso: http://locus.ufv.br/handle/123456789/4815

Regardless the kind of method used to evaluate an experiment of sugarcane families, it remains the question of the number of necessary individuals to represent the performance of these families. The evaluation of the ideal sample size to represent with precision and efficiency the mean, variance and other parameters in families is necessary and has been study, contributing to the optimization of resources available in a breeding program. The literature shows certain divergence and lack of consistency on results obtained for the best sample size in sugarcane. In this context, the aim of this study was to determine the minimum number of plants that should be sampled in experiments of sugarcane families, for yield traits, which would allow an effective estimation of some parameters used to assist the family selection. Data from an experiment carried out at Sugarcane Breeding Center (CECA) at Universidade Federal de Viçosa, in the municipal district of Oratórios MG, were utilized. The experiment was planned as a randomized complete blocks desing with 18 full-sib families of sugarcane in six repetitions and individual information of 20 plants per plot. It was carried out a study of resampling with replacement, followed by an estimation of some genetic and phenotipic parameters. Also the Modified Maximum Curvature Method and Sampling Intensity Method were used. The estimates of sample size varied accordingly to the parameter being estimated, to the trait being evaluated and to the method being utilized. The resampling method allows for a more efficient comparison of the sample size effects on the estimation of the genetic and phenotipic parameters, when compared to the Modified Maximum Curvature and Sampling Intensity Methods. A sample of 16 plants per family would be sufficient to obtain good estimates of all the parameters discussed, for all the traits considered. However, in the circunstance of a separate sampling by trait, a sample of 10 plants per family would allow an efficient estimation of the parameters for Brix.