出版社：Grupo de Pesquisa Metodologias em Ensino e Aprendizagem em Ciências
摘要：Estimates of the repeatability coefficient have been used in studies with different species of palm trees, helping to define the appropriate number of measurements to be collected. The objective of this work was to determine the minimum number of evaluations necessary for plant and bunch traits in areas of natural occurrence of the species of bacaba O. distichus and O. bacaba, for an efficient collection process. The study was developed using data from expeditions to collect these species carried out in five municipalities in the State of Pará. For O. bacaba, data from the provenances Baião, Marabá and Terra Santa were used, while for O. distichus, data were from the provenances Belém, Baião, Marabá and São João do Araguaia. The following plant and bunch traits were measured: stem circumference at breast height, length of five internodes, total bunch weight, number of rachillae per bunch, length of the central rachis, weight of one hundred fruits. Data were submitted to four different repeatability methods. The repeatability coefficients presented different magnitudes based on the methods applied, for the traits studied in O. bacaba and O. distichus. The method of the principal components resulted in more adequate estimates. For an efficient field collection process of these species of bacaba, eleven measurements in O.bacaba and four in O. distichus are considered as the minimum number to reach a level of 90% of determination, in areas of spontaneous occurrence.
