Longitudinal analysis of Pinus caribaea var. hondurensis full-sibling progenies based on multivariate analysis

Abstract

Aim of study: To define an early selection strategy based on tests applied to full-sibling progenies of Pinus caribaea var. hondurensis grown in the Cerrado Biome. Area of study: Prata region (MG), Brazil. Material and methods: Progeny tests were cultivated in 2006; the study followed a completely randomized design, with 79 families of full-siblings and 15 repetitions, with one plant per plot. Thinning was carried out at the age of 6 and 8 years; 615 individuals and 44 families were included in the test. The following quantitative variables were used in the statistical analysis of data on the remaining individuals: diameter at breast height (DBH) in cm, total height (H) in m, and volume in dm³ at the age of 3, 4, 5, 6, 7, 8 and 11 years. BLUP multi-trait multivariate model, with non-structured covariance structure matrix, was adopted for calculations. Main results: There were strong additive genetic correlations (above 90%) between variables DBH and H, in all analyzed ages. Strong volume correlations were estimated based on the age group over four years; volume selection efficiency reached its peak at the age of five years. Selection based on volume at the age of 5 years leads to genetic gains in this variable; selection intensity values can range from 7.8% to 6.4% and 5.4%, and from 10% to 20% and 30%. Research highlights: The best strategy lies on carrying out the selections at the age of five years, based on 30% selection intensity.

Keywords

• early selection
• multi-trait BLUP
• unstructured matrices