Parâmetros genéticos em teste de procedências e progênies de Toona ciliata M. Roemer var. australis

Abstract

The increase the dimensions of the trees, by genetic improvement, without considering their architectures can make them mechanically unstable in situations of high winds. As consequence may occur falls, breaks and permanent buckling of stems, causing economic losses and environmental damage. In the present study had as objective to estimate genetic parameters and predict wood production gains with improvements in the trees architecture. It was used a test of provenances/progenies/plants of Toona ciliata M. Roemer var. australis, established in Campo Belo, MG, with 74 months old. This test involved 12 provenances of seeds widely distributed in the Australian east coast. Data of diameter to breast height (DBH), total tree height (Ht), volume (Vol), slenderness coefficient (S = Ht.DBH -1 ) and stem form (For), at 31, 54 and 74 months of age, there were obtained in individuals of 63 progenies from all provenances. The genetic parameters and prediction of genotypic values were estimated by REML/BLUP method. There was great genetic variability among and within provenances, showing that the test has an important role for the improvement and conservation of the species ex situ. Whereas the criterion of having a higher frequency of trees with low S or equal to 0.75, established by the average value of the population, it was found that DBH of the trees with 54 months of age was the most suitable characteristic for the selection, given the large size effective of the population selected. Thus, in this age 54 months were estimated heritability values in the narrow sense of 0.34, 0.37, 0.31, 0.14 and 0.29, respectively, for the characteristics DBH, Ht, Vol, S and For. Genetic correlations among DBH x Ht, DBH x Vol and Ht x Vol, were 0.92, 0.96 and 0.88, respectively. The genetic correlations between the growth characteristics and stem architecture as: DBH x S, DBH x For, Ht x S and Ht x For were, respectively -0.51, 0.69, -0.20 and 0.68. For related gains, the DBH was the best predictor feature to improve the wood production and the S. On the other hand, the selection by height as the characteristic gain predictor correlated should be avoided, since the timber volume can be increased, but without occurring desirable reductions of S and thus, improving the mechanical stability of the trees.