02059naa a2200169 a 450000100080000000500110000800800410001910000110006024500520007126000090012350000200013252016150015265300230176765300150179065300200180577300640182510757312011-04-15       2008    bl uuuu     u00u1 u    #d1 aEpagri  aEvaluation of Training Systems for Nashi Pears.  c2008  aISSN, 0567-7572  aIn Brazil, Nashi pears are being cultivated by small growers looking for high quality fruits. This fruit crop requires special training care due its fragile branches that break easily during the growing season. The traditional training system is the so called trellis, which is very expensive and is restricting the crop expansion. In 1996, an experimental orchard with three varieties of Nashi pears was planted to evaluate their behavior regarding to three training systems and pruning. Central leader, vase, and V training systems were tested on Hosui, Kosui and Nijisseiki pears grafted on Pyrus calleryana rootstok, spaced 6 x 6 m. The experimental design was RCB, with three replicates. The training systems were set in the plots and the varieties in the sub-plots. Tree vigor by measuring Trunk Cross Sectional Area (TCSA), yield and mean fruit weight were evaluated. After 8 years, ?Hosui? was the best adapted. ?Kosui? was too susceptible to the branches canker caused by Botryosphaeria sp., with more than 50% of dead plants. ?Nijisseiki? had the lowest yield independently of the training system. The central leader and the V training systems were those with the highest yield, independently of the varieties. Yields were generally low due to the high rate of floral abortion. The training systems did not affect the mean fruit weight. The spacing used was to large for all the training systems tested, suggesting that it can be reduced. The central leader system showed to be viable as a free system, without support for the trees, which reduces significantly the installation costs of the orchard.  aOrchard management  aPera nashi  aPyrus pyrifolia  tActa Horticulturae, Leuvengv. 800, n. 2, p. 755-762, 2008.