Título:	Bayesian and frequentist statistical analysis in quantitative genetics applied to fores trees breeding in Brazil.
Autoria:	RESENDE, M. D. V. de FERNANDES, J. S. C. BERTOLUCCI, F. DE L. G DUDA, L. L.
Afiliação:	MARCOS DEON VILELA DE RESENDE, CNPF.
Ano de publicação:	2000
Referência:	In: FOREST GENETICS FOR THE NEXT MILLENNIUM, 2000, Durban. Proceedings... Scottsville: Institute for Commercial Forestry Research, 2000.
Conteúdo:	This paper describes some efforts made in Brazil in the framework of breeding (additive genetic) and genotypic (additive + dominance) values prediction and variance components estimation in a tree breeding context. Emphasis is given to the use of the mixed linear models techniques including both, frequentist and Bayesian approaches. For growth traits, which are associated to growth trajectories or curves through the ages, alternative models like repeatability, multivariate and random regression models were'compared. Using data from a Eucalyptus urophy!ta progeny test, evaluated at several ages (1,2......7 years), the random regression or covariance functions model has performed successfully. The application (using data from unbalanced diallel mating designs of Pinus caribaea var. hondurensis) of Bayesian techniques, implemented trough stochastic simulation (Gibbs sampling), showed that additional results in relation to frequentist approach are obtained.
Thesagro:	Pinus Caribaea
Palavras-chave:	Random regression Bayesian statistics Mixed model equations Individual BLUP REML
Notas:	IUFRO Working Party 2.08.01. Tropical Species Breeding and Genetic Resources.
Tipo do material:	Artigo em anais e proceedings
Acesso:	openAccess
Aparece nas coleções:	Resumo em anais de congresso (CNPF)