Introduction

Konkan region of Maharashtra (India) is well known for the commercial production of world famous Alphonso variety of mango, which presently occupies an area of more than 0.1 million ha, spreading along the western coast of the state. Due to excellent fruit quality and tang of the fruit, it is in great demand both in domestic and international markets. Although, the area under mango cultivation has approximately doubled during the last decade, the yield per unit area has decreased by 14 % (FAO 2008). The productivity of Alphonso in Konkan is very low (2.5 t ha−1) as compared to a national average (6.5 t ha−1) and has been a long standing constraint (DBSKKV 2014). Among the several factors ascribed for low productivity, alternate bearing habit is considered as first and foremost reason. Many studies have reported irregularity of flowering in mango, which varied with time and intensity of flowering from year to year to almost complete biennial (alternate) flowering habit (Singh 1972; Shiva and Usha 2014).

High levels of endogenous gibberellins in cv. Alphonso, warm and humid climatic conditions of the coastal Konkan region have been found to favour excessive and repeated vegetative flushing at the cost of flowering (Kachru et al. 1971; Tomer 1984; Burondkar and Gunjate 1993). Under such circumstances, growth retarding chemicals of triazoles group, e.g., paclobutrazol, that can inhibit gibberellins biosynthesis, has been used by researchers to correct the biennial bearing habit (Burondkar and Gunjate 1993; Burondkar et al. 2000; Murti and Upreti 2000; Nartvaranant et al. 2000; Blaikie et al. 2004; Yeshitela et al. 2004; Nafees et al. 2010). Paclobutrazol application has been recommended two to three weeks before flowering for inducing regular flowering with an increase in yield of Alphonso. Each year, approximately 20 kl of PBZ is used for induction of regular flowering in mango. However, the effect of PBZ on internal hormonal composition in mango has not been studied under Konkan climatic conditions. Therefore, the present experiment was conducted to study the effect of PBZ on the hormonal composition of cv. Alphonso under Konkan condition.

Materials and methods

Plant material and treatments

A study was conducted during 2011 in the mango orchard (latitude 17°74′N and longitude 73°18′E) with 12 years old uniform grafts of cv. Alphonso planted at 10 m spacing. Soil of the experimental site was sandy loam and average canopy diameter of the trees was 8.0 m. Recommended dose of PBZ was applied once as soil drench at a concentration 0.75 ml a.i. m−1 of canopy diameter by spreading the solution in a circular band of 25 cm width around the tree trunk during third week of August. Three trees were assigned both under paclobutrazol (PBZ+) and control (PBZ−) treatments and each tree was regarded as one replication.

Sampling and data collection

Hundred terminal shoots measuring average length of 20 cm were tagged in four directions of the experimental trees. Observations on the response of PBZ on flower induction were monitored at 15 days interval after 1 month of PBZ application. Periodical sampling at weekly interval was made from 2nd week of September till visual indications of bud burst were evident. The analysis of phytohormones in leaves and terminal buds in the treated and control trees were made at 15 days before bud break, at bud break stage and at 15 days after floral bud induction stage.

Indole acetic acid analysis

Indole acetic acid (IAA) was assayed by Salkowsky’s colorimetric method (Mayer 1958). The presence of IAA in samples was analyzed by thin layer chromatography (TLC) and high performance liquid chromatography (HPLC) using technical grade IAA as standard. Thin layer chromatography chromatograms were run on thin silica gel plates. Solvent system used to separate indole compounds was ethyl acetate: butanol: ethanol: water [5:3:1:1 (v/v)]. Indole acetic acid was detected on TLC plates by spraying with Kovac’s reagent (Tien et al. 1979). Indole acetic acid assay was done by injecting 20 μl of the filtered extract onto a five μm reverse phase column (Waters Associates μBondapak C18, 250 mm × 4 mm) in a waters associates HPLC equipped with an ultraviolet detector absorbing at 208 nm. The solvent system used to separate indole compounds was water: acetonitrile [76:24 (v/v)], and the flow rate was maintained at 1 ml min−1. Quantification was done by comparison of peak heights.

Abscisic acid and cytokinins analysis

The bud and leaf samples (2 g fresh weight) were macerated in chilled 80 % methanol and filtered. The filtrate was evaporated in vacuum at 35 °C, the residue dissolved in water and the pH was adjusted to 3.0. The aqueous acidic extract was partitioned twice against equal volume of chilled diethyl ether. The ether phase was separated and dried over anhydrous sodium sulfate. The extract was kept overnight at 4 °C, filtered and after evaporating ether; the residue was dissolved in 20 mM Tris-buffer, pH 7.8 for the ABA analysis. For the cytokinins analysis, the pH of the aqueous extract left after ether partitioning was adjusted to 8.0 and partitioned thrice with equal volumes of water-saturated n-butanol. The butanol phase was separated, evaporated in vacuum at 35 °C and the residue was dissolved in 20 mM Tris-buffer, pH 7.8. The ABA (Weiler 1982) and the cytokinins—ZR and DHZR (Barthe and Stewart 1985) content were determined by ELISA using an ELISA reader (Multiskan MS 352, Labsystem, Finland), by employing laboratory raised polyclonal antibodies.

Gibberellins analysis

Gibberellins assay was done using gas chromatography mass spectrometry (GC–MS). Samples for gibberellin analysis using SIM (selected ion monitoring) were prepared according to Kaufman et al. (1976). Samples of apical buds (2 g) were separately grounded in liquid nitrogen and extracted thrice with chilled 80 % aqueous methanol (8 ml g−1). Pooled filtrates were dried in vacuum, dissolved in 10 ml of water and pH adjusted to 3.0. The aqueous acidic extract was partitioned against ethyl acetate three times, and ethyl acetate phase after separation was evaporated, re-dissolved in 0.1 M phosphate buffer, pH 8.0 and stirred with 2 g of insoluble polyvinylpyrrolidone. After filtration, pH was lowered to 3.0 and acid extract was partitioned thrice with ethyl acetate. After evaporating ethyl acetate, the gibberellin samples were purified on charcoal-Celite (1:2 w/w, 5 g charcoal) adsorption column and silicic acid 100 mesh partition columns (Powell and Tautvydas 1967). Gibberellins were eluted from charcoal-Celite column employing 80 % acetone and gradiently eluted using 30–65 % ethyl acetate in n-hexane from silicic column. The purified gibberellins were passed through 0.45 µm Millipore filter (Fluoropore). For gibberellins assay by GC–MS, the samples were first methylated and further derivatized to trimethylsilyl (TMS) ether with 50 µl of bis-(trimethylsilyl) trifluoracetamide (BSTFA) + 1 % trimethylchlorosilane (TMCS) at 60 °C for 1 h. The derivatized samples were analysed on GC–MS system (Varian, USA) equipped with 3800 series gas chromatograph and 4000 series ion trap mass spectrometer. A VH-5 column (100–200 mesh, 30 m × 0.2 mm id) programmed at 100  °C for 7 min, followed by a linear increase @ 10 up to 200 °C with 3 min hold and @ 5 °C with 19 min hold (total run time of 55 min) was employed to resolve silylated gibberellins. The split ratio was 1:10. The helium gas inlet was pneumatic pressure controlled at a constant flow of 1 ml min−1 and injector, transfer line, source and trap temperatures were maintained at 285, 270, 210 and 200 °C, respectively. The data were obtained in the full scan mode. Identification was confirmed on the basis of retention time and co-occurrence of additional ions employing NIST-2007 library. The quantification was done on the basis of peak area and the standard gibberellins (Sigma-Aldrich, USA).

Results and discussion

Effect of paclobutrazol on indol acetic acid and abscisic acid content

It was observed that, IAA content in the PBZ treated trees during flowering period was less than control trees. Moreover, PBZ treatment lead to decrease in IAA content in leaf and bud before bud break and at bud break stage. Increment was observed in the IAA content in leaf and bud 15 days after bud break stage (Table 1). Interruption in IAA biosynthesis leads to a strong increase in root derived cytokinins (Bangerth et al. 2000). It may be concluded that PBZ treatment induced decrease in IAA may be due to inhibition of IAA biosynthesis, which subsequently increased cytokinins production in roots, finally leading to increase in its level in the apical buds.

Table 1 Effect of paclobutrazol on indol acetic acid, abscisic acid and cytokinins content in leaves and buds of cv. Alphonso (data represent means; n = 3)
Full size table

In the control trees ABA content ranged from 453.23 to 608.22 and 325.05–411.90 ng g−1 in the buds and leaves, respectively. Following PBZ application, the ABA content increased by 15.98–58.51 and 53.93–65.69 % in buds and leaves, respectively and the increase was greater at the bud break stage. With respect to the stages, the ABA content was highest in buds at 15 days after bud break stage in the control trees and at bud break stage in the PBZ treated trees. In leaves, ABA was highest at bud break stage in treated and untreated trees (Table 1). The role of high ABA content in the induction of floral bud formation has been reported by Chacko (1986) in mango and by Hou et al. (1987) in litchi. Paclobutrazol induced increase in ABA is in accordance with the earlier reports of induction in ABA biosynthesis as an upshot of the altered isoprenoid pathway, which is partially common in ABA and gibberellin biosynthesis (Murti et al. 2001; Abdel-Rahim et al. 2011). Increase in ABA could also be a consequence of reduced ABA catabolism to phaseic acid by triazole compounds as reported by Hauser et al. (1990). The high ABA level is expected to induce bud dormancy, which consequently help in floral bud formation as flowering in mango is found to occur on the buds at rest.

Effect of paclobutrazol on fractions of cytokinin

The results of the present study revealed that, zeatin riboside (ZR) was prominent cytokinin in the buds of control trees (Table 1). The total of cytokinin contents increased significantly in the buds from 15 days before bud break stage till bud break stage and decreased after 15 days of bud break (Table 1). Leaf cytokinin content was marginally influenced by PBZ application. Both in the control and treated trees the trends with regard to different stages were similar, with cytokinin contents attaining peak at the bud break stage. The gain in the amount of cytokinin in the buds at the bud break stage and induction in cytokinin contents by PBZ gave strong evidence of a role of cytokinin in floral induction in mango. In mango, Chen (1987), Abdel-Rahim et al. (2008) and Upreti et al. (2013) reported increase in cytokinin at floral bud formation stage. Increase in cytokinin by PBZ could either be due to stimulated synthesis of different cytokinins that are transported to shoots or arresting of cytokinins degradation (Grossmann 1992). Cytokinins are suggested to act positively in floral bud induction by regulating cell division process. Further, the involvement of cytokinins in floral bud formation has also been reported by Chen (1985) through accelerated floral bud formation in mango by external application of benzyl adenine.

Effect of paclobutrazol on fractions of gibberellin

Among the various types of gibberellins, GA3, followed by GA1, GA4, GA5 and GA7 were the major gibberellins identified in the buds of control trees (Table 2). Other minor gibberellins identified were GA9, GA13 and GA25. In the control trees, gibberellin contents showed increasing trends from 15 days before bud break stage till bud break stage, while at 15 days after bud break its content slightly decreased. Following PBZ application, various gibberellins contents decreased in the buds of trees. In treated trees, total gibberellin contents was highest at 15 days before bud break stage and decreased drastically at the time of bud break stage. In treated trees slight increase in gibberellin contents was observed at 15 days after bud break stage (Table 2). Furthermore, the magnitude of decline in major identified gibberellins by PBZ was dependent on stages. Paclobutrazol induced decline in GA3 was more at bud break stage. Paclobutrazol induced decline in various gibberellins is substantiated by the earlier reports that PBZ inhibits gibberellin biosynthesis (Fletcher et al. 2010). Reduction in gibberellin contents favouring floral bud formation has earlier been reported by Chen (1987), Abdel-Rahim et al. (2008) and Upreti et al. (2013) in mango, Chen (1990) in lychee and Koshita et al. (1999) in citrus, and it may be related to a reduction in internodal length or the accumulation of carbohydrates favoured by declining gibberellin contents.

Table 2 Effects of paclobutrazol on different fractions of gibberellins (GA) inbudof cv. Alphonso (data represent means; n = 3)
Full size table

Decline in the gibberellin content is an important contributor to floral bud initiation, which has also been reported by the studies of Tomer (1984) and Oosthuyse (1996). Further, it was noticed that, not all gibberellins have similar role to play in the induction of floral buds. Decline in GA1 content may be important in preparing the buds for floral induction, decline in GA3, GA4 and GA7 may act in floral bud initiation. It is important to note that the side chain in the cytokinin molecule is obtained from the isoprenoid pathway responsible for gibberellin biosynthesis, thus PBZ by blocking gibberellin biosynthesis may channelize the metabolic intermediates to cytokinins production.

The similarity in the trends of cytokinins and gibberellins in the control and treated trees with respect to stages revealed a role of these hormones in the formation of floral buds in mango. It is also important to note that increase in ABA and cytokinin contents with decline in IAA and gibberellins in buds in response to PBZ application are important attributes contributing to floral bud production in mango.

Conclusion

The results of the present study showed that application of PBZ as a soil drench affected the hormonal composition of the plants. Application of PBZ increased total ABA and cytoknin contents, and decreased IAA and gibberellins contents, especially GA1, GA3, GA4 and GA7 in the shoot, which resulted in enhanced flowering and reduced the extra flush of vegetative growth during the reproductive stage of mango plants. Among the fractions of gibberellin, major fraction was GA3, and in case of cytokinins major fraction was zeatin riboside (ZR) in mango cv. Alphonso.