Viability of premature selection of Pinus taeda L. in diameter breast height in genetic improvement program

The superior genotypes selection of Pinus taeda L. using genetic field trials require many years, due to the rotation age of commercial plantations, which consumes a lot of financial resources. Regarding this, many researchers have put their efforts in evaluate the selection in early ages prior to the final age of rotation. The goals of this study are to estimate, at the early age of eight years and final rotation age of fifteen years, the genetic parameters and the predicted genetic gains between and within the progenies of Pinus taeda L. half-sib family, with experiments established in three distinct locations in the state of Parana, Brazil. There were significant correlations observed for the progenies ranking and for the linear correlation of the additive genetic gain between the age of eight and fifteen years. The obtained heritability in this study were greater in Campo do Tenente, were it was observed a greater growth, also concurring with the location of the evaluated half-sib family selection. For the Campo do Tenente experiment, the eight years old measures displayed the same predicted genetic gain of those from the fifteen years old measures, for either the parents or the best individual selection. For the Morungava experiment, located in the municipality of Senges, having similar soil and climate conditions of Campo do Tenente, however further away from the parents selection location, it was only possible to succeed in the best individual selection at eight years old age, whereas the parents selection at this age have not showed significant statistical difference apart the control. For the best individual selection at early age, it is recommended to utilize the maximum number of individuals per genitor restriction, due to showing the same statistical results of the selection at fifteen years old, but with greater genetic variability. For the Mocambo experiment, also in the municipality of Senges, but with soil and climate conditions very different, it was not possible to detect statistical differences for neither the parents nor the best individual selection.