Performance of Avocado (Persea americana Mill.) and Mango (Mangifera indica L.) Seedlings Compared with Their Grafted Trees

Most fruit-tree breeding projects are based on selection of seedlings in regard to their performance. The selected seedlings are vegetatively propagated, usually by grafting. It is highly important for the breeder to know whether the performance of the grafted tree will resemble the performance of the original seedling. In this study the performance of avocado and mango seedlings was compared with that of their grafted duplicates. Significant differences were found in only 8 out of 36 avocado traits and 2 out of 10 mango traits. Significant seedling x graft interaction was detected in 10 other avocado traits. These differences were considered of no practical significance, since their magnitude was of minor importance for the breeder. The conclusion for avocado and mango breeders is that for most traits selection could be carried out on ungrafted seedlings. In most fruit tree breeding projects, the testing phase is based on vegetative propagation of selected seedlings, usually by grafting. It is known though, that the performance of a grafted tree results from the interaction of the rootstock and the scion (Ben-Ya'acov, 1987; Rom and Carlson, 1987). Rootstocks affect the adaptability of the tree to various soil conditions such as drought, flooding, salinity, and alkalinity (Rom and Carlson, 1987). Scion vigor, manifested principally in tree size, is well known to be affected by rootstock x scion interaction. Since the fifteenth century, the use of rootstocks to dwarf trees has become a common practice, and it has been widely used Received for publication 29 March 1994. Accepted for publication 21 July 1994. Contribution No. 1042-E, 1993 series, from the Agricultural Research Organization, The Volcani Center, Bet Dagan, Israel. We thank the staff of the Akko and Bsor Experiment Stations and especially D. Zamet and R. Rotem for their excellent technical assistance. This project was financed in part by the Israel Fruit Board. The cost of publishing this paper was defrayed in part by the payment of page charges. Under postal regulations, this paper therefore must be hereby marked advertisement solely to indicate this fact. in this century (Tukey, 1964). A dwarfing rootstock tends to reduce seasonal shoot extension of the scion. Such trees are characterized by heavy production at an early age. This is exemplified in the case of apple trees grafted on East Malling IX rootstock, which reach only about one third of the normal size and produce fruit at an early age. Rootstocks may also affect ripening time, fruit color, and fruit size (Rom and Carlson, 1987; Wutscher, 1979). Rootstocks are of great importance in the modern intensive culture of avocado. The practice of avocado grafting started in Florida at the beginning of this century (Popenoe, 1939). Avocado rootstocks have been selected in California for their relative tolerance to phytophthora root rot (Coffey, 1987) and in Israel for their higher tolerance to salinity (Kadman and Ben Ya'acov, 1976) and calcareous soils (Ben-Ya'acov et al., 1979). Avocado rootstocks have been reported to influence tree size and productivity. Tree size has been found to be affected by rootstock in California (Bergh and Whitsell, 1962). In a large-scale research project on rootstock-scion relationship conducted in Israel, rootstocks have been found to significantly affect avocado tree size and productivity (Ben-Ya'acov 1972; Ben-Ya'acov et al., 1979). Tomer and Rotem (1989) found 'Degania 400' seedling rootstocks to produce larger trees with higher production than 'Nahlat 3'. Rootstocks may have also influenced avocado productivity; trees budded on 'Nabal' seedlings have given appreciably higher yields than those budded on four other rootstocks (Oppenheimer, 1960). As for mango, Oppenheimer (1958) compared three polyembryonic rootstocks in Israel and found 'Sabre' to be superior in growth and production to 'Warburg' and '14.12'. In a second trial trees grafted on two other rootstocks ('14.6' and ‘14.7') were equal to 'Sabre' in yields, while trees grafted on '3.2' and '14.12' were less productive (Oppenheimer, 1968). Giri and Yacub (1965) reported that polyembryonic mango rootstocks delivered more vigor to the scions, compared with monoembryonic rootstocks. Swamy et al. (1972) reported that mango vigor and yield are not always controlled by the rootstock, while Sen (1939) and Janhati (1972) suggested that the scion (and not the rootstock) is responsible for the tree shape. Most fruit-tree breeding projects utilize the selection of the best performing non-juvenile seedlings, concerning agriculturally important traits. These seedlings are then vegetatively propagated, usually by grafting, to allow a more thorough assessment. Since the first-stage selection is carried out on nongrafted seedlings, it is highly important for the breeder to be aware of any potential effect on performance caused by either the rootstock or the rootstock x scion interaction. Obviously, in cases where such an effect is expected, much caution is needed at the seedling selection stage. This study was conducted to determine the potential effect of grafting on the performance of avocado and mango seedlings. Materials and Methods Assessment of avocado seedlings and their grafted trees was carried out at the Akko Experiment Station in western Galilee, Israel (Lavi et al., 1990). Seedlings were planted in breeding plots at distances of 4m between double rows of 2 x 1m. The juvenile period was shortened by the use of autumn girdling (Lahavet al., 1986). Seedlings with outstanding fruit characteristics and productivity were grafted in the nursery and also topworked on Mexican and/or West-Indian seedling rootstocks. Each of the selected seedlings was grafted on two to four rootstock types (one to four seedlings each). Seventy original seedlings and their grafted duplicates were assessed for thirty-six traits (Table 1). Productivity, one of the economically most important traits, was compared in a larger population of 229 graft-seedling pairs. These included the above mentioned 70 seedlings which were grafted on young rootstocks in the nursery and also topworked grafts (one to three trees each). The other 159 seedlings were grafted in only one of the two methods. Evaluation of productivity (in five grades) was averaged for all the grafted trees of the same seedling. The same trait in the 70 original seedlings and their grafted duplicates was evaluated in 4 degrees of evaluation only (see fruit density on the tree in Table 1). Comparison between mango seedlings and their grafted trees was conducted at the Bsor Experiment Station in the southwest of Israel (Lavi et al., 1989). Seedlings were planted in a breeding plot at distances of 4 x 2 m. Seven interesting seedlings were selected and each was grafted on seven '13/1' polyembryonic seedling rootstocks (Gazit and Kadman, 1980). A comparison of the performance of the original seedling with its grafted duplicates for 10 traits was carried out 4 years after grafting. Seedlings and grafted trees of avocado and mango were located in the same orchard, under similar soil and climatic conditions. Similar agrotechniques were performed for the seedlings and for their grafted duplicates. However, the grafted duplicates were usually prepared only after seedlings were selected, at the age of 6 to 10 years. Thus, as a rule, evaluation of the original seedlings and their grafted duplicates was performed in different years. Only once did we have the opportunity to compare the performance of seedlings and their grafted duplicates at the same period of time. In that case, graftwood was taken from 12 seedlings in the nursery and seedlings and their grafted duplicates were planted, side by side, at the selection plot. Evaluation of avocado and mango seedlings and grafted duplicates was based either on measurements or on visual scoring (Tables 1 and 2). Two-way analysis of variance was carried out for each trait between seedlings and their grafts by Program GLM of SAS (SAS Institute, 1989).