Application of transmissible agents to control citrus tree size.

Abstract

Establishing citrus orchards at higher densities has become standard practice in South Africa. These plantings make more efficient use of inputs such as land, fertilizer, irrigation and spraying, and they are capable of high early production and economic returns. However, disadvantages develop over time, depending on tree density and climate. These include overcrowding between and in rows, which induces difficult access by vehicles, poor insect control in dense canopies, and a decrease in yield and quality due to dieback and overcrowding. It is thus important to find a method to control tree size to maintain the benefits of such plantings. All three main tree size controlling methods viz. the use of genetic dwarf plant material, management practises and the use of transmissible agents are currently under investigation in the country. The aim of this study was to identify transmissible agents that can be used successfully to control tree size without detrimental effects. Research was done to identify transmissible agents as well as their application as dwarfing agents for commercial use. An important safety measure that should be considered before a transmissible agent is introduced for tree size control of citrus, is that such an agent should be endemic in the industry. Citrus tristeza virus (CTV) is a well known endemic disease in Southern Africa and virus-free shoot-tip grafted (STG) material is pre-immunised with avirulent isolates to reduce its effect. STG is also used to remove citrus viroids (CVd's) and the use of such agents as dwarfing agents is in contradiction with the aims of the Southern African Citrus Improvement Programme. However, research over many years in Australia has shown that avirulent CVd's can be used succesfully to control tree size. It is thus important to identify suitable CVd isolates for commercial application but it is also important to know which CVd's occur in commercial orchards. The occurrence of CVd's in old commercial citrus orchards, planted prior to the introduction of shoottip grafted material, was established by biological indexing with Etrog citron (Citrus medica L. cv. Arizona-861-S-1). Twenty-one percent of the trees indexed tested positive for CVd's. Sequential polyacrylamide electrophoresis (sPAGE) was employed to establish which viroids are most commonly present. Only two groups of CVd were detected, viz. citrus exocortis viroid (CEVd) and CVd-III. The latter CVd was the most common (40%) while CEVd occurred in only 8% of the samples. More than a third of the samples were infected with both CVd's. Two CVd isolates, a mild and a severe isolate according to Etrog citron reactions, were bud-inoculated to Delta Valencia on rough lemon, Poncirus trifoliata and four trifoliate hybrid rootstocks. Growth, production and occurrence of disease symptoms of these trees were compared to trees on the same rootstocks without CVd. All the trees were pre-immunized with the standard tristeza virus isolate. Results of sPAGE analysis of nucleic extracts indicated that the severe isolate (CD 11) contained CEVd only, and the mild isolate (CD 12) contained CVd-III. Overall, both isolates caused a significant reduction in tree size. The cumulative production over five years of the CD 11 infected trees did not differ from the CVd-free trees although the trees were smaller. This was due to a significantly higher production efficiency (kg/m(3) canopy). The production efficiency of the CD 12 trees was similar to the CVd-free trees, but the smaller trees resulted in a significantly lower production. Disease symptoms occurred with both isolates, but symptoms differed. Poncirus trifoliata var. Rubidoux was more sensitive to CVd isolates than four trifoliate hybrid rootstocks. Marsh grapefruit trees on Troyer citrange rootstock were bud-inoculated with different CTV isolates prior to planting in the field. Selected CTV isolates GFMS 2, GFMS 10, GFMS 12, GFMS 19, GFMS 25, GFMS 27 and GFMS 35, free of citrus viroids, were bud-inoculated into the virus-free plants. A severe isolate (GFSS 1) and plants that were left virus-free served as controls. Tree size, production, fruit size and tree health were determined. Fifteen years after planting, canopy volumes of trees with three isolates, GFMS 2, GFMS 19 and GFMS 25, were significantly smaller than the control trees as well as trees with isolates GFMS 10, GFMS 12 and GFMS 35. Trees with isolate GFMS 19 had a larger diameter than those with isolates GFMS 2 and GFMS 25. Together with a slightly higher yield efficiency, GFMS 19 trees resulted in a cumulative yield equal to that of the control and the GFMS 12 trees. Considering fruit size and their value, the performance of trees with isolate GFMS 19 equalled that of the larger trees. Tree health was also similar which makes this isolate suitable for use in high density plantings. A projection was made which showed that the production and crop value of to trees with isolate GFMS 19 were similar to those of trees with isolates GFMS 10, GFMS 12 and GFMS 35. However, benefits such as easier and better spray application and easier harvesting can increase profits when trees with isolate GFMS 19 are planted. The dwarfing characteristics of four isolates, CD 4, CD 8, CD 9 and CD 10, derived from healthy looking dwarfed field trees were evaluated. They were bud-inoculated to Delta Valencia trees on Yuma citrange rootstock prior to planting in the field. Five years after planting, isolates CD 4 and CD 9 successfully reduced canopy volumes by 60%, and CD 10 by 30%, without any detrimental effects. No CVd's could be detected biologically or by sPAGE in these three isolates. Isolate CD 8 however, contained two viroids, CEVd and CVd-III, but had no deleterious effects on the rough lemon rootstock. CTV was the only other pathogen present in the isolates. Indexing for cachexia, psorosis, impietratura and tatter leaf was negative. The dwarfing abilities of the isolates are therefore attributed to isolates of citrus tristeza virus. Production was according to tree size and the yield efficiency of the inoculated trees was equal to that of the uninoculated control trees. External and internal fruit quality was not affected. The trees were naturally infected with huanglongbing (greening) five years after planting, but the disease remained low for several years in trees with isolate CD 4. Three transmissible isolates (CD 4, CD 9, CD 10), derived from dwarfed field trees, were compared with two CVd isolates (CD 8, CD 12) for their abilities to control tree size of sweet orange. The isolates were bud-inoculated to Valencia on rough lemon, two Poncirus trifoliata and three trifoliate hybrid rootstocks, and compared to uninoculated trees on the same rootstocks. Isolates CD 4, CD 9 and CD 10 gave no reaction on Etrog citron and sPAGE of nucleic acid extracts failed to detect known CVd's. The two CVd isolates gave severe and mild reactions on Etrog citron and sPAGE showed that CD8 contained CEVd and CVd-III while CD 12 contained only CVd-III. The effect of the isolates on tree size, production, production efficiency and disease occurrence were monitored. Overall, CD 4, CD 9 and CD 10 did not reduce canopy volumes while trees with CD 8 and CD 12 were significantly smaller. However, CD 4 significantly reduced canopy volumes where Yuma citrange was used as a rootstock. In contrast, CD 8, containing CEVd, did not reduce canopy volumes on this rootstock, while CD 12 reduced it significantly. None of the trees on the other rootstocks were affected by either CD 4, CD 9 or CD 10. Canopy volume reductions by CD 8 and CD 12 differed from each other with all the trifoliate rootstocks. The production efficiency of trees with the two CVd isolates was significantly higher than the control trees as well as those with CD 4, CD 9 and CD 10. The higher efficiency of these trees resulted in cumulative production equal to the uninoculated trees. Disease symptoms occurred where all the isolates were inoculated, however, symptoms as well as the susceptibility of the rootstocks, differed among each other. Delta Valencia trees on Yuma citrange rootstock were inoculated respectively with two mild (GFMS 12 and T55), an intermediate (GFMS 10) and a severe (GFSS 1) CTV isolate prior to planting in the field. The same CTV isolates were also inoculated in combination with a CVd- III isolate. A virus-free control was included in the trial. All the CTV isolates were without the Seedling Yellows component of CTV. Seven years after planting, canopy volumes of the trees with the two mild isolates and the control were smaller than those of trees with the intermediate and severe isolates. This was in contradiction to what was expected. Overall, the CVd isolate had an additional reducing effect on canopy size, but only those trees with mild isolate T55 and severe isolate GFSS 1 were significantly affected. Production on a per tree basis was according to the canopy sizes, thus, the trees carrying the intermediate CTV isolate were the highest producers. The production efficiency (kg/m(3) canopy), however, did not differ among trees with the CTV isolates and the control. The CVd isolate generally increased the production efficiency. The internal quality of the fruit was not affected by any treatment. The lack of a suitable genetic dwarfing rootstock for citrus makes it essential to evaluate alternative methods to reduce tree size for high density plantings. Four transmissible dwarfing factors, derived from dwarfed trees, were evaluated for commercial application in a hot (Malelane) and intermediate (Nelspruit) production area. Generally, trees in the hot production area were more vigorous with a lower production efficiency than trees in the intermediate area. The dwarfing effects of isolates CD 4, CD 9 and to a lesser extent, CD 10, were reduced in the hot area. Isolate CD 8 caused no dwarfing at either location. The reduction of dwarfing at the hot site may be attributed to the suppression of CTV by high temperatures. Currently some of the isolates which were tested in this investigation are applied on a larger scale in different climatic areas as commercial trials. Formal trials are continuing and they are aimed to elucidate the dwarfing characteristics as well as the inducement of disease of the four viroids in the CVd-III group.