SELEÇÃO DE PROGÊNIES DE Eucalyptus pilularis Smith POR MATRIZ DE PARENTESCO OBTIDA POR MEIO DE MICROSSATÉLITES

Abstract

Meeting the global demand for wood and forest by-products has been one of the main challenges today and in the coming decades. Among the raw materials, the species Eucalyptus pilularis stands out because it has good wood quality, a high growth rate, and the ability to develop in different types of soils. Due to the characteristics of the species, there is great potential for resuming breeding programs, considering the genetic diversity of E. ilularis materials existing in Brazil. Based on this information, this research aimed to evaluate the potential of microsatellite markers to assist in the selection of E. pilularis parents in the formation of a new population for reproduction and to analyze the genetic variability through quantitative variables of progenies. The experiment was installed in a randomized block design (DBC), with five replications, in linear plots consisting of five plants, spaced 3 m x 2 m. In this experiment, 13 provenances and 115 progenies were evaluated. The analysis of the quantitative traits was carried out using a mixed model, using a matrix of half-sibling relationships and the molecular relationship of the population, obtained through the coancestry coefficient, using microsatellite markers. The genetic parameters were estimated by the methodology (REML/BLUP), for the characters diameter at breast height (DBH, cm), total height (H, m), and volume (Vol, cm³) at 5 years of age in Itatinga, SP. The genetic gain of the population was also estimated taking into account its productivity in volume, considering the molecular parentage matrix. The population of E. pilularis presented a survival rate of 85.45%, with mean values of H of 18.65 m, DBH of 14.28 cm, and Vol of 14.57 cm³. For the parentage matrix considering half-siblings, the individual heritability (ℎ𝑖 2) for the evaluated characters was considered low to moderate (0.14 to 0.20), while for the average heritability (ℎ𝑚2) they were considered high magnitude (0.53 to 0.61). These results were similar for the parentage matrix estimated by markers, with an individual heritability (ℎ𝑖2) considered of moderate magnitude (0.27 to 0.40). By estimating the predicted genotypic values considering the molecular parentage matrix, for the character volume (cm³), the half-sibling matrix and the molecular parentage matrix presented low to moderate accuracy (𝑟𝑎𝑎̂ ), respectively, 0 .37 to 0.45 and 0.52 to 0.63. Considering selection for genetic conservation, the genetic gain was 28.40%, with an improved population with an average volume of 18.71 cm³. A mass selection simulation was also carried out with an intensity of 10% of the best E. pilularis individuals. In the 10% selection simulation, all provenances remained in the experiment, but 28 progenies did not present any superior individual to be maintained in the experiment. In this selection, the genetic gain was 30.93%, with an average volume of the improved population of 19.08 cm³. The use of the molecular kinship matrix provided a better adjustment in the estimates of genetic parameters, increasing their estimates and enabling more realistic results for the species. The volume character presented a relative coefficient of variation close to 1, in addition to presenting higher values for the evaluated parameters compared to the other evaluated characters (height and DBH), and its use is recommended for selecting the best individuals. To create an improved population aiming at greater genetic gain and effective size, one must opt for mass selection, whereas if the objective is to create a base population, one must opt for selection aiming at genetic conservation, that is, leaving one individual from each progeny in each block, this way the best individuals from all origins and progenies will be part of this population, thus guaranteeing greater genetic diversity in the improved population.