YIELD AND PROFITABILITY OF HIGH-DENSITY PEAR PRODUCTION WITH PYRUS ROOTSTOCKS

A 2-ha field trial was planted in 2003 at Geneva, New York State, United States which compared four training systems/densities [Central Leader (598 trees ha^-1), Vertical Axis (1281 trees ha^-1), Tall Spindle (2243 trees ha^-1) and Super Spindle (5382 trees ha^-1)] on 6 rootstocks (seedling, OHF97, OHF87, Pyrodwarf, Pyro 2-33 and Quince A) with 2 cultivars (‘Bartlett’ and ‘Bosc’). Tree density had a large effect on cumulative yield while rootstock genotype had a much smaller effect. With ‘Bartlett’ the super high-density Super Spindle system had 4 times the cumulative yield as the low-density central leader system while with ‘Bosc’ the medium density Tall Spindle system had twice the yield as the low density central leader system. There was little effect of rootstock with ‘Bosc’ but with ‘Bartlett’, OHF97 and OHF87 had the highest cumulative yield at each density. Quince A had the lowest yield at all densities. Pyrodwarf and Pyro 2-33 had similar yields at all densities but Pyrodwarf had numerous root suckers. Fruit size was negatively related to planting density with the super spindle system producing significantly smaller fruit size than other systems. Part of the effect was due to greater crop loads on the super spindle system. However, when fruit size was adjusted for crop load there was still a negative effect of planting density on fruit size of ‘Bartlett’. With ‘Bartlett’, Quince A produced the largest fruit size while Pyrodwarf and seedling had significantly smaller size. With ‘Bosc’ there was no effect of rootstock on fruit size. Yield efficiency was not affected by rootstock genotype but there was a slight negative relationship between planting density and yield efficiency. Profitability of the four orchard systems was estimated as the extra income after 11 years of the 3 high-density systems compared to the traditional low-density pear orchard of 600 trees ha-1. With ‘Bartlett’, income was highest for the Super Spindle system and lowest for the central leader system. The relationship of density and income was curvilinear with little difference between the Tall Spindle system at 2,200 trees ha^-1 and the Super Spindle system at 5,500 trees ha^-1. With ‘Bosc’ the relationship was quadratic with the highest net extra income with the Tall Spindle system and there appeared to be an optimum density at about 1,900 trees ha^-1 with significantly lower income at 5000 trees ha^-1.