Hindawi Publishing Corporation ISRN Forestry Volume 2013, Article ID 747591, 7 pages http://dx.doi.org/10.1155/2013/747591 Research Article Genetic Structure of a Loblolly Pine Breeding Population at Brazil Juliane Rezende Mercer, 1 Milena de Luna Alves Lima, 2 Antonio Rioyei Higa, 2 Chirlei Glienke, 1 and Marina Isabel Mateus de Almeida 1 1 Departamento de Gen´ etica, Universidade Federal do Paran´ a (UFPR), Setor de Ciˆ encias Biol´ ogicas, 81531-980 Curitiba, PR, Brazil 2 Setor de Ciˆ encias Agr´ arias, Universidade Federal do Paran´ a (UFPR), Curso de Engenharia Florestal, Avenida Professor Lothario Meissner, 900, Jardim Botanico, 80210-170 Curitiba, PR, Brazil Correspondence should be addressed to Juliane Rezende Mercer; juju.mercer@hotmail.com Received 1 April 2013; Accepted 23 April 2013 Academic Editors: G. Martinez Pastur, P. Newton, and H. Zeng Copyright © 2013 Juliane Rezende Mercer et al. Tis is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Te genetic structure of a Brazilian loblolly pine (Pinus taeda L.) breeding population, represented by 120 open-pollinated families, was determined using Bayesian inference and genotypes of 15 microsatellite (simple sequence repeat (SSR)) loci in 1,130 seedling progeny. Te 120 maternal parents had been phenotypically selected about 15 years ago for wood volume in fve diferent forestry plantations (FPs) in the south of Brazil. Additional selection for wood volume, based on a previous progeny test, was applied to the frst best (i) and second best (ii) tree per block within each family. We adopted a procedure of “learning samples” to fnd the most likely number of inferred genetic clusters () or ancestral populations. Te frst hypothesis that was rejected was that the most probable value of =5 was coincident with the fve FPs, since the FPs were, a priori, assumed to be from 5 diferent backgrounds or origins. It was used the familiar structure of the population to infer the genotypes of maternal ancestors. It was concluded that the maternal generation is the most likely to have been planted by the mixture of three diferent seed sources or origins, that there are fve genetic groups (=5) in the population of progeny, and that they have been formed from the occurrence of assortative mating and also from a strong pressure in the selection within families. Te trees with the best genetic value (i) maintained a higher genetic variability when compared to the trees of second best performance (ii), with higher values of heterozygosity and of numbers of maternal alleles that were kept the same. Te migration model that best explains the results is the contact zone model. Te population diferentiation ( ST ) was 2-3 times higher in ofspring than in relation to the maternal generation. Te relevancy of the results and the way they were explored may be of value both for studies of population genetics, as for plant breeding programs, since they help monitoring the population’s genetic variability during generations of selection. 1. Introduction Loblolly pine (Pinus taeda L.) is a monoecious conifer, diploid tree with a predominant cross-fertilization. Its native range is in southeastern United States, and it was frst introduced in Brazil in the 1940s. Its wood is used for industry panels and sawn timber. It is a fast-growing tree and is well adapted to temperate climate that makes it a good candidate for planting forests in southern Brazil. Among 5.74 million hectares (ha) of all existing forest plantations in Brazil in 2006, P. taeda occupies 1.52 million hectares in the southern states [1]. Most of these plantations were established from geneti- cally improved seeds produced in clonal seed orchards that were established by breeding programs using phenotypically selected trees. Te main objective of this research was to study the genetic structure of a breeding population that consists of open-pollinated progenies of multiple P. taeda families using microsatellite or simple sequence repeats (SSR) markers and a Bayesian clustering approach implemented in the STRUCTURE sofware (Pritchard et al. [2]) to infer groups or subpopulations in this population.