Relação entre estatísticas do planejamento experimental e tamanho de parcela para pimentão em estufa plástica

• Ano de defesa: 2008
• Autor(a) principal: Lorentz, Leandro Homrich
• Tipo de documento: Tese
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Santa Maria; BR; Agronomia; UFSM; Programa de Pós-Graduação em Agronomia
• Palavras-chave em Português: Capsicum annuum; Precisão experimental; Tamanho da parcela; Ambiente protegido; Experimental precision; Plot size; Greenhouse; CNPQ::CIENCIAS AGRARIAS::AGRONOMIA
• Link de acesso: http://repositorio.ufsm.br/handle/1/3170

The interior of a plastic greenhouse whose plants had been cultivated on soil can be considered a heterogeneous environment due high variability in the production, being necessary the use of experimental techniques for improvement of the statistical inferences precision. Between the main forms of obtain experimental error reduction is the use of the optimal plot size that has special importance in experiments inside plastic greenhouse, where the experimental area is normally restricted to the preexisting installations. Therefore, it is necessary to know the optimal plot size (Xo) for the cultures as well as understanding the relations among the parameters that determine it. The aim of this work is to estimate the relation of the equations parameters that determine the optimal plot size and estimate the optimal plot size for fruit fresh phytomass of chili pepper cultivated in plastic greenhouse. It was conducted four blank experiments under plastic covering, each one with ten culture lines of 70 plants. The fruit fresh phytomass was evaluated in each plant, identifying its position inside the greenhouse, by the line number, and its position in the line in each harvest. The optimal plot size was estimated and related with the parameters used for its estimate by the Pearson correlation coefficient, the path analysis and the response surface methodology. In the end of this work is recommended the use of parcels with ten plants in the line and at least three harvests for parcel. It was observed that Xo was negatively correlated with the number of harvested plants and the average of each culture line in the harvests. The positive and high correlation of the coefficient of variation estimate for the parcels of a basic unit and Xo indicates that this is its main determinative one, reducing the importance of the heterogeneity index (b), that if related weakly with Xo.