Performance of cornell-geneva rootstocks across north america in multi-location nc-140 rootstock trials

From 1992 to 1999, five multi-site replicated rootstock trials were established by the US national rootstock testing group, "NC-140". The trials compared elite Geneva apple rootstocks which were bred for tolerance to fire blight and Phytophthora root rot, high yield efficiency and good tree survival to commercial standards. The named Geneva rootstocks are designated as Geneva or "G" stocks while unnamed numbered selections from Geneva are designated as 'CG' stocks. In the 1992 plots which used 'Liberty' as the scion, Geneva 11 and CG.3029 had the highest cumulative yield efficiency, good tree survival and also had good average fruit size. They had similar tree size as M.9, but exceeded the yield performance of M.9. In contrast, Geneva 65 was more dwarfing than M.9 and had significantly lower cumulative yield efficiency and smaller fruit size than M.9. Among semi-dwarf stocks, Geneva 30, CG.6210, CG.4222 and CG.5179 all exceeded the performance of M.7 and MM.106. In the 1993 plots which also used 'Liberty' as the scion, CG.4247, CG.3041, CG.3902 and CG.3007 had the highest yield efficiencies and had good tree survival. All were similar in size to M.9, but performed significantly better than M.9 or M.26. Among the semi-dwarf stocks top performers were G.30, CG.6210, CG.222 and Geneva 202. All performed significantly better than M.7. In the 1994 plots which used Gala as the scion, Geneva 30 produced a tree similar in size to M.26 and more efficient than M.26 at 12 sites and less efficient at 6 sites. Fruit size was similar to M.26. G.30 generally had good survival; however, in 5 of 23 sites 50-60% of the trees broke off at the graft union during wind storms. In the 1998 and 1999 trials, Geneva 16 has been slightly larger than M.9 with Gala, Fuji and McIntosh, but similar to M.9 with Jonagold. Productivity of Geneva 16 has been similar to M.9 in all trials. CG.3041 has been similar in size and productivity to M.9 with Jonagold, Fuji and McIntosh. INTRODUCTION The Cornell University apple rootstock breeding project located at Geneva NY has the objective of developing rootstock genotypes with improved nursery and orchard characteristics and that are better adapted to the biotic stresses of fire blight (Erwinia amylovora), and crown rot (Phytophthora spp.) which are common in New York state and surrounding areas (Cummins and Aldwinckle, 1983). Progeny from planned crosses underwent greenhouse screening procedures to select for tolerance to fire blight and crown rot. Surviving genotypes were then tested for their value as rootstocks by evaluating their propagation characteristics in the nursery, and productivity and dwarfing in the orchards at the New York State Agricultural Experiment Station in Geneva. By the Proc I IS Rootstocks – Decid. Fruit Eds. M.A. Moreno Sanchez and A.D. Webster Acta Hort 658, ISHS 2004 242 early 1990's, Jim Cummins had selected a group of promising fire blight resistant apple rootstocks with a range of dwarfing. These were then tested in second level trials that included on-farm trials within NY state (Robinson and Hoying, 2003) and national trials in the US, Canada, France (Masseron and Simard, 2002) and New Zealand (personal communication from Stuart Tustin). In this paper, we report on results from the national US and Canadian trials conducted by the NC-140 rootstock research group. MATERIALS AND METHODS Since 1992, the elite selections from the Geneva breeding program have been planted into several of the multi-location national rootstock trials conducted by the NC140 project to assess productivity, precocity, survival and field tolerance to abiotic and biotic stresses. In 1992, 1993, 1994, 1998 and 1999 replicated rootstock trials were planted at multiple locations in the United States and Canada (Table 1). Each trial had from 1-7 CG rootstock clones with appropriate Malling rootstock controls and other rootstocks of interest to the NC-140 group. Each site had from 6-10 single tree replications of each rootstock. For each trial all of the plant material was grown in a common nursery. The plots had a uniform tree spacing and trees were also managed in a uniform manner. However, certain orchard practices such as fertilization, irrigation and thinning were left to local control. Several sites were terminated after 5 years, but the rest were continued through year 9 or 10. The trials planted in 1992 and 1993 used 'Liberty' as the scion while the 1994 trial used 'Gala' as the scion. The 1998 trial had 'Gala' and 'Jonagold' as the scions while the 1999 trial had 'McIntosh' and 'Fuji' as the scions. RESULTS In the 1992 plots with 'Liberty' as the scion, G.11 had the highest cumulative yield efficiency along with good tree survival and good average fruit size (Table 2). Its tree size was similar to M.9, but it had better yield efficiency than M.9. In contrast, G.65 was more dwarfing than M.9 and had significantly lower cumulative yield efficiency and smaller fruit size than M.9. G.202 was not significantly different than M.9, but was ranked between M.9 and M.7 in size and yield efficiency. Among the semi-dwarf stocks, G.30 and CG.6210 were top performers with significantly greater yield efficiency than M.7, MM.106 and MM.111. G.30 generally had good survival; however, in one site several trees broke off at the graft union. CG.6210 had good survival at all sites. In the 1993 plots with 'Liberty' as the scion, CG.4247, CG.3041, and CG.3007 had the highest yield efficiencies and had good tree survival (Table 3). All were smaller than M.26 and performed significantly better than M.26. In this trial, G.65 was significantly larger than the 1992 trial. It had similar yield efficiency as M.26. Among the semi-dwarf stocks top performers were G.30, and G.202. They were similar in size, but G.30 had significantly greater yield efficiency. Both performed significantly better than M.7. In the 1994 plots with 'Gala' as the scion, G.30 was similar in size to M.26 at most of the 25 sites. At 7 sites it was smaller in size than M.26 and at 8 sites it was larger than M.26. Among the sites where G.30 was larger than M.26 it appeared that at several of the sites, particularly western North America sites, M.26 was stunted while G.30 was not. G.30 had higher yield efficiency than M.26 (Table 4). The yield efficiency of G.30 was similar to M.9. G.30 had significantly more root sucker than either M.9 or M.26. Although G.30 has excellent tree yield performance it has had the lowest tree survival. At several sites a significant number of trees of Gala on G.30 broke off at the graft union in years 4-7 during high winds. In the 1998 plots with Jonagold, G.16 and CG.3041 were similar in size to M.9, but with Gala, G.16 was significantly larger than M.9 (Table 5). Although the trees were larger, the yield efficiency of Gala/ G.16 was similar to M.9. The trees in the Gala plot were derived from tissue cultured liners which may have been the reason for the increased tree vigor with Gala/ G.16 compared to Gala/M.9. In the 1999 plot G.16 was similar in size to M.26, but CG.3041 was smaller in size than M.26 (Table 6). Both had higher yield efficiency than M.26. CG.5935 was larger in