Avaliação genética de linhagens de poedeiras utilizando componentes principais e função discriminante linear de Fisher

• Autor(a) principal: Michelotti, Vanessa Tomazetti
• Data de Publicação: 2018
• Tipo de documento: Dissertação
• Idioma: por
• Título da fonte: Manancial - Repositório Digital da UFSM
• dARK ID: ark:/26339/0013000014zrw
• Texto Completo: http://repositorio.ufsm.br/handle/1/14403
• topic: Análise de variância multivariada; Correlação de posição de Spearman; Herdabilidade; Seleção; Taxa de produção de ovos; Heritability; Multivariate analysis of variance; Rate of egg production; Selection; Spearman's rank correlation; CNPQ::CIENCIAS AGRARIAS::ZOOTECNIA

The objective of this work was to evaluate genetically the individual performance characteristics of laying hens, egg quality and rates of egg production from the 19th to the 70th week of age of lines of hens of the Rhode Island Red (GG and MM) and Plymouth Rock White (SS). The data set used came from the Centro Nacional de Pesquisa de Suínos e Aves of the Empresa Brasileira de Pesquisa Agropecuária (CNPSA/EMBRAPA). The genetic values and estimates of heritabilities for each trait within each lineage were obtained through a univariate animal model. The characteristics that explain most of the genetic variation of this database were identified using Principal Component analysis, Spearman's position correlation, and later Fisher's Discriminant Functions were created. Through the principal components analysis, the quality characteristics that explained most of the genetic variation of the data were the density (D36), weight (PO36) and length x width ratio (R36) of the egg measured at the 36th week of production. The production characteristics that best represented the total production rate were: the production rates accumulated at the 50th (TA50) and 60th (TA60) weeks and the partial production rate from the 23rd to the 40th week (TP23a40). By obtaining the Fisher's discriminant functions (FDFs) for both sexes in the three lines studied, it was possible to observe that the productive characteristics (posture rates) were the most important in the composition of the Function. Higher correlations were observed between FDF1 and FDF3 (ranging from 0.97 to 0.99); followed by FDF1 and FDF2 (ranging from 0.86 to 0.94). A selection of 20% of the genetically superior animals of the last generation was performed, and the coincidence of selected animals in the different FDFs with FDF1 was presented as a percentage. It can be observed that the FDF3 function selected 100% of males in common with those selected by FDF1, in the MM and SS lines. The FDF3 presented greater coincidence in all lineages and sexes with FDF1. Thus, the variables identified as most representative are D36, PO36 and R36 along with TA60, and the Fisher discriminant function that considers all these characteristics is efficient to anticipate the selection of the animals.