Evaluation of Graft Compatibility for Taxonomic Relationships among Species of the Orange Subfamily

Callus tissues induced from shoots of species in the orange subfamily (Aurantioideae) were grafted in vitro to evaluate the correlation between reaction at the callus graft interface and taxonomic relationship. The graft interface between two pieces of callus from species that are taxonomically close was not distinguishable by anatomical observations. On the other hand, in the combinations of less closely related species, the border of graft interface was distinct. In combinations of species that were even more distant taxonomically, the border was clear and some deposits accumulated. In these combinations the decay of cell wall was evident at the contact surface of both species of callus cells. The response of the grafted callus cells and whether or not incompatibility was evident, was correlated with the taxonomic relationships of the species. INTRODUCTION Grafting is the art of connecting two pieces of living plant tissue together so that they will unite and subsequently grow and develop as one composite plant. Many fruit and nut trees have been propagated by grafting. A number of detailed studies have been made on graft union formation with woody plants (Hartmann et al., 1997). In general, grafting compatibility increased with the closeness of taxonomic relationship (Roberts, 1949). However, there are many exceptions to this rule (Andrews and Marquez, 1993; Hartmann et al., 1997). The distinction between a compatible and incompatible graft union is not always clear-cut. An incompatible graft is not synonymous with an unsuccessful graft. Thus, incompatibility should be used in a restricted sense and will refer only to the mutual physiological influence between tissues of the stock and scion that ultimately result in an unsuccessful graft union (Moor and Walker, 1981). It is difficult to carry out a uniform experiment on grafting compatibility, because the success of grafting is affected by the art and craftsmanship in grafting. Fujii and Nito (1972) demonstrated that the application of the tissue culture system was useful to examine the events at the initial stage of graft union formation. They observed the recognition of callus tissue between graft partners during the early stage of grafting. It is postulated that some of the factors that affect whether a graft is compatible or incompatible occur when the callus cells first touch (Yeomen et al., 1984). In commercial application to fruit trees, the grafted partners usually belong to the same species or genus. The compatible and incompatible features in grafting can provide information on relative taxonomic and metabolic similarities. Callus cultures for graft compatibility studies can offer several advantages. The environmental variables and the art of practical grafting are reduced, and changes in development can be assessed easier and faster. In this experiment, the behavior of in vitro grown callus cells during the initial stages of grafting were described to evaluate the taxonomic relationship between species within the orange subfamily. Citrus, Poncirus and Fortunella are all classified in the ‘True Citrus Fruit Trees’ group (Swingle and Reece, 1967). They are sexually compatible, producing hybrids (Swingle and Reece, 1967; Iwamasa et al., 1988). Trifoliate orange (Poncirus trifoliata) is used commercially as a rootstock for Citrus and Fortunella cultivars. The genera Murraya, Triphasia, Atalantia and Citropsis are more distantly related. 85 Proc. IInd IS on Biotech. of Trop & Subtrop. Species Eds: W.-C. Chang and R. Drew Acta Hort 692, ISHS 2005 MATERIALS AND METHODS Seeds were collected from mature fruit of Murraya koenigii, Triphasia trifolia, Atalantia monophylla, Citropsis schweinfurthii, Fortunella polyandra (Malay Kumquat), Poncirus trifoliata (‘Normal’ type and ‘Flying Dragon’) and Citrus natsudaidai (‘Tachibana Orange’). These species have been maintained in the germplasm collection of species of the orange subfamily at Saga University. Seeds were aseptically germinated on the agar medium for one month. Calli were induced from internodes of seedlings on 1/2 Murashige and Tucker (1969) (MT) medium supplemented with 1 μM kinetin, 3 μM 2, 4-D and 3 μM GA3 for one month. The calli removed from explants were proliferated on MT medium containing 20 μM kinetin and 6 μM 2, 4-D for 4 weeks with two passages of subculture to fresh medium at two week intervals. Cultures were incubated at a temperature of 26 ± 2 °C and under 35 μmol m s light intensity with 16/8 hours photoperiod of light / dark. Callus was cut into small pieces of 4 5 mm wide and surfaces to be placed in contact were freshly cut. The different combinations were established by placing two callus pieces together on MT, with 20 μM kinetin and 6 μM 2,4-D. To maintain a close contact between the two callus tissues, calli were enclosed by a glass ring 16 mm in diameter and incubated under the same condition used for callus culture. The combinations were sampled 2 and 4 weeks after culture. The callus of Citrus natsudaidai was combined with the other 6 species and the homograft was used as a control. To evaluate the contact interface, parts of the graft were cut out and immediately fixed with 99% ethanol and acetic acid (3:1 v/v) and embedded in paraffin after dehydration in a graded ethanol series and clearing in xylene. Specimens were sectioned at 20 μm sections. Sections were double stained with 1% ethanol solution of safranin and 1% fast green. Cells at the graft interface were observed under light microscopy and photographed. RESULTS AND DISCUSSION Two pieces of callus cultured in a glass ring continued growing and contacted each other after 2 weeks of culture. After 4 weeks, they pushed one another to form the graft union (Fig. 1). Even the combinations of two pieces of callus from different genera were normal in appearance. The graft interface between two pieces of callus from homografts or taxonomically close species was not clear anatomically (Fig. 2). In the homograft of Citrus natsudaidai, the border between two pieces of callus was not clear and cells were closely adherent (Fig. 2A). In the combination of two strains of Poncirus, i.e. ‘Normal’ and ‘Flying Dragon’, the border was distinguishable because of the different size of callus cells, but there was no necrotic or isolation layer between them (Fig. 2B). Although C. natsudaidai is in a different genus from Poncirus trifoliata and Fortunella polyandra, their graft interfaces were not distinguishable (Fig. 2C and D), supporting the grouping of these three genera into the ‘True Citrus Fruit Trees’ group (Swingle and Reece, 1967), and the graft compatibility between Poncirus trifoliata and Citrus and Fortunella cultivars. It is obvious that these three genera are genetically close. On the other hand, in the combinations that are taxonomically further apart, the border of graft interface was clear (Table 1, Fig. 3). In these combinations, some deposits were accumulated at the interface between two pieces of callus. Especially, in the combination of C. natsudaidai and Atalantia monophylla, the border was densely stained by safranin, suggesting the presence of killed cells (Fig. 3B). The biochemical nature of substances which accumulated to form the border is unknown. The presence or absence of a border between callus of C. natsudaidai and other genera grafted in vitro is listed in Table 1. The callus combinations between different genera that are outside the ‘True Citrus Fruit Trees’ group, formed clear borders between the two pieces of callus. Cell differentiation to form xylem was not observed in the callus after 4 weeks culture. Other factors would be required for the differentiation of vascular system in tissue