Comparative financial efficiency of training systems and rootstocks for 'Alpine' nectarines (Prunus persica var. nectarine)
Most nectarine orchards in South Africa are currently planted at a distance of 4 x 1.5 m (2 500 trees/ha). These trees are mainly sylleptically trained to a central leader, although many producers also use the proleptic route. The former produces relatively high yields early in the lifetime of the orchard. A problem with nectarine production in South Africa is the lack of efficient rootstocks in terms of aspects such as size-control and the use of nematode-resistant rootstocks. The aim of this study is to evaluate different training systems for nectarine production and to investigate the role of three rootstocks that play a dominant role in the peach industry in South Africa. ‘Alpine’ nectarines were planted in the winter of 2002 at Lushof near Ceres, Western Cape, South Africa (33º18’S, 19º20’E). The trees were trained according to four different training systems: a four-leader system (5 x 3 m; 667 trees/ha), a two-leader system (5 x 1.5 m; 1 333 trees/ha), a proleptically trained central leader (5 x 1 m; 2 000 trees/ha), and a sylleptically trained central leader (5 x 1 m; 2 000 trees/ha). The trees were planted on three different rootstocks: GF 667; SAPO 778; Kakamas seedling. The time spent per tree on pruning, thinning and picking was recorded. During harvest, the number of fruit and fruit mass per tree were recorded. Light measurements were recorded annually after summer pruning. The measurements were taken at different heights and at different depths in the canopy. To compare the training systems on an economic basis, the data from the trial together with projected data gathered from farmers and advisors were used to calculate the net present value (NPV) and internal rate of return (IRR) for each training system. The results showed that rootstock only played a significant role when it came to fruit mass (fruit size). Fruit from trees on SAPO 778 were heavier, indicating bigger fruit, than fruit from trees on Kakamas seedling rootstocks and this can play a role in packout percentage and income. In terms of the training system, the four-leader system took the most time to manage per tree. However, this system took the least time to manage per hectare during the initial years. No differences were found between the two central leaders. They both took the longest time to manage per hectare. The four-leader system produced significantly less fruit than any of the other systems during the first two years of production. In the third year of production, there was no significant difference found between the systems. Light penetration seemed to be the poorest at the middle and bottom of the canopy for trees trained to a central leader. Because of the open centre of the four-leader system, light penetration into the middle of these trees was good, but poor light penetration occurred in the upper and outer parts of the canopy underneath the scaffold branches. Poor light penetration occurred in the parts lower than 1.5 m from the ground for all the systems. This was the area that was measured in this study. The result of an economic comparison showed that according to the IRR rating, the fourleader system should be preferred. The final decision should however be made according to the NPV rating. Results obtained from NPV calculations did not lead to the same conclusions as could be made from the IRR calculations. According to the rating of the NPV at five percent discounting rate, the two-leader should be the preferred system, while the proleptically trained central leader system should be preferred at a ten percent discounting rate. This implies that when the opportunity cost is low, the two-leader system should be preferred, and when the opportunity cost is high, the proleptically trained central leader system should be preferred.