SELECTION METHODS TO OPTIMIZE THE GAIN AND GENETIC DIVERSITY IN Pinus caribaea var. caribaea

The proposal of this work was to estimate the genetic variability in orchards of Pinus caribaea var. caribaea based on growth traits and to analyze the best selection method. This study was conducted in two areas of P. caribaea var. caribaea situated in Savannah biome. The first orchard was a randomized complete block design with 76 progenies and 4 controls (area 1), the second orchard, the lattice design was 10x10 with 99 progenies and one control (area 2), 28 and 27 years old, respectively. The software SELEGEN was used to estimate genetic parameters trough REML/BLUP method. Significant variation was observed between and with progeny all traits in area 2 and only between plants within plots for height in area 1. The highest estimates of genetic variation and heritability were obtained for area 1. Without the optimization of selection, the highest gain (4.8%) in the selection between and within with a selection intensity of 52%, for area 1. In area 2, the highest gain (2.86%) in individual selection. We conclude that there is low genetic variability in seedlings orchards of P. caribaea var. caribaea. However, area 1 presents higher genetic control than area 2, and should be better explored. For the next generations, it is recommended the infusion of new genetic material to proceed with a forest improvement program, since it was observed low variability and low gains in the selection of P. caribaea var. caribaea.