Alternativa para ajuste de estande em testes de progênies e clonais na cultura do eucalipto

Abstract

Efficiency in the evaluation of eucalyptus progenies and/or clones depends on some factors, which involve: the experimental design choice; the number of environments; data analysis tools; and especially the rigor in conducting and getting the experimental data. In this context, plants loss, in experimental plots is one of the main problems of all plant breeders. For perennial crops, this problem is even more serious due to neighboring plant longer exposure time to one or more failures, also because the selective unit in progeny tests is the individual. Thus, the objective of this research was: to estimate the effect of failures or dominated plants in progenies or clones tests conducted with the eucalyptus crop; verifying if the available area per plant is a good strategy to mitigate the effect of failures; evaluate if the effect of failures varies when using plots with single or multiple plants; and finally to verify if the association between the performance of the plant in the progeny test (PT) and its clones in the clonal test (CT) is affected by the existence of failures or dominated plants in the experiments. Five PTs and four CTs were conducted in the state of São Paulo by the company FibriaCelulose S.A. Of the PTs, three were progenies of full sibs and two of half sibs. PT plants were evaluated at five years and CTs at three years. The evaluated characteristic was the circumference at breast height (CBH) in cm, from which the diameter at breast height (DBH) was estimated. In order to carry out the adjustment, the area available to the plants, neighboring the failures, was used. The dominated plants were also considered in the adjustment, with the DBH one or two standard deviations below the mean. It was found that: in experiments with failure levels lower than 20% the adjustment using the area available to the plants did not improve the selection efficiency in progeny and clonal tests, regardless of whether the experiment is carried out with one or several plants per plot; the strategy of considering in adjustment not only the failures, but also the dominated plants, is not advantageous especially when we consider as dominated trees with DBH a standard deviation below the mean; both the use of the BLUP estimates and the individuals phenotype in the PTs were of low efficiency in the identification of the superior clones in the CTs, with or without the adjustment of the failures.