Interaction of training system and rootstock on yield, fruit size, fruit quality and crop value of three sweet cherry cultivars

A 2002 field trial compared 'Lapins', 'Regina' and 'Hudson' cherry cultivars on three Gisela (Gi.5, Gi.6, Gi.12) and Mazzard seedling rootstocks each planted in four training systems (Quad Axis - 598 trees ha(-1), Spanish Bush - 748 trees ha(-1), Central Leader - 748 trees ha(-1), and Vertical Axis - 997 trees ha(-1)). After 11 years, tree size on Gi. 5 was the smallest, followed by trees on Gi. 6, G. 12 and Mazzard, respectively. Trees on all three Gisela stocks were much more precocious than Mazzard. Cumulative yield after 11 years was highest for trees on Gi. 5 followed by Gi. 12, Gi. 6, and Mazzard. Average fruit size was largest on Gi. 12. With the self-fertile cultivar 'Lapins', Gi. 5 and Gi. 6 induced excessive production, which resulted in smaller fruit size. However, with the self-infertile cultivar, 'Regina', Gi. 5 was the only rootstock to induce commercially acceptable high yields while also producing the largest fruit size. Although Gi. 12 was a vigorous tree, it had much greater precocity and productivity than Mazzard. Among training systems, the Vertical Axis system had the highest cumulative yield ha(-1), cumulative yield efficiency and cumulative crop value followed by Central Leader, Spanish Bush and Quad Axis. Cumulative yields and crop value were positively related to planting density. Fruit size was largest with the Quad Axis and the Central Leader, smaller with the Spanish Bush, and smallest with the Vertical Axis. Fruit size was negatively related to planting density.