2019 Volume 88 Issue 1 Pages 92-99
Flowering induction in Amorphophallus muelleri Blume is required for mass production of seeds within a short period of time. Hence, the objective of this study was to evaluate the effect of gibberellic acid (GA3) application on flowering induction for non-flowering-aged corms (immature corms) in A. muelleri. Research was conducted at Leuwikopo Farm, Bogor-Indonesia in May 2015–August 2016. Dormant corms of bulbils (0 year), 1-, and 2-year-old were treated with 0, 500, 1000, 1500 and 2000 ppm GA3. A 3-year-old corm was used as an additional control. Results demonstrated that GA3 application effectively induced flowering in 2-year-old corms. Bulbils flowered at a rate of 1.7–6.7% and 1-year-old corm at a rate of 2.8–16.7% irrespective of GA3 levels; 100% and 50–100% of inflorescences, respectively, were abnormal leading to low seed production. The GA3 application caused erratic flowering in bulbils and one-year-old corms, indicating age-dependent-flowering in A. muelleri. The erratic flowering caused a detrimental effect on seed production. In this experiment, 2-year-old corms produced larger inflorescences and a lower number of abnormal inflorescences. Thus, application of GA3 at a level of 1500 ppm on 2-year-old corms is recommended for mass production of A. muelleri seeds within a short period of time. In the future, it will be interesting to study the physiological mechanism of the erratic flowering phenomenon across corm age in A. muelleri.
The cultivation area of Amorphophallus muelleri Blume (locally called iles-iles or porang) is increasing in tropical Asia; because its corm is a source of low-cost glucomannan (KGM) (Zhao et al., 2010). The KGM flour is widely used as a functional food (Zhang et al., 2005; Alonso-Sande et al., 2009). The plant is suitable for use in agroforestry systems (Sugiyama and Santosa, 2008), leading to steady expansion of its cultivation in regions without land competition with existing cash crops (Santosa, 2014).
The A. muelleri is traditionally propagated by aerial bulbils and triploid apomictic seeds for which setting occurs without pollination (Dani, 2008; Jansen et al., 1996). The use of seeds as a propagation material is increasing in A. muelleri production because it makes it possible to harvest uniform-sized corms earlier than the use of bulbils and corm sections (Zhang et al., 2010). The importance of seed production also arises in the context of genetic studies including speeding up breeding of new variety development. Nevertheless, commercial seed production in A. muelleri is cumbersome because it takes 38–48 months from seed planting to seed harvest (Sumarwoto, 2005; Sugiyama and Santosa, 2008).
Although studies on flower induction in A. muelleri and related species have been conducted (Itoh, 1960; Latifah and Purwantoro, 2015; Santosa et al., 2006; Zhao et al., 2010, 2013), the studies indicated the morphological appearance of reproduction organs and seed quality. For example, in A. paeoniifolius that constitutes starch in its corm, Santosa et al. (2006) found that application of GA3 stimulated flowering on young seed corms but mostly produced abnormal flowers followed by low seed set. Thus, the GA3 effect on young seed corms in A. muelleri is that mainly constituted of glucomannan is less understood. The objective of this study was to evaluate the GA3 application on flowering induction at different corm ages in order to clarify and find appropriate condition to enhance seed production of A. muelleri.
Experiments were conducted on virgin Latosolic soil (Darmaga series) at Leuwikopo Farm IPB, Bogor Indonesia (260 m above sea level) from May 2015–August 2016. One month prior to planting, one ton·ha−1 of lime (CaCO3) was applied to all plots to raise the soil pH to 5.6. After lime application, the soil contained a low amount of total N (0.28% by Kjeldhl method), a low amount of Bray I phosphorus (34.0 ppm) and a medium level of available potassium (175 ppm). Soil comprised 12.2% sand, 24.1% silt and 63.7% clay. Two weeks prior to planting, goat manure of about 0.5 kg per corm (equal to 10 ton·ha−1) was put into in each planting hole. The manure (pH 7.8) contained 1.13% of total N, 0.07% of phosphorus and 0.28% of potassium. The average temperature during experiment was 26.3°C (23–32°C) with 80–84% air relative humidity. Light intensity was maintained by using an artificial shading net to reduce light intensity to 65% of full sunshine. Watering was done daily.
Planting materials, bulbils and corm, were harvested at the dormant stage in May, 2015 (dry season). After harvest, bulbils and corms were stored at room temperature until GA3 treatment. Corms from different ages, i.e., bulbils, 1-year-old, and 2-year-old corms were treated with GA3 at levels of 0 (control), 500 ppm (0.5 g·L−1), 1000 ppm (1.0 g·L−1), 1500 ppm (1.5 g·L−1) and 2000 ppm (2.0 g·L−1), on August 2, 2015. In addition, 3-year-old corms were maintained as a control. Bulbils were considered to be 0-year-old corm, and corm developed from the bulbils was considered as 1-year-old corm. A. muelleri bulbils took two and three growing seasons to produce 2-year-old and 3-year-old corms, respectively. At time of treatment, the average weight (mean ± SD) was 21.6 ± 3.2 g, 186.7 ± 9.5 g, 721.6 ± 79.1 g and 1199.3 ± 245.5 g for bulbils, 1-, 2-, and 3-year-old corms, respectively; bud size was 0.3 to 0.5 cm for 1-, 2- and 3-year-old corms and no visible buds were found on bulbils.
A certain amount of GA3 powder (BioReagent, Thomas Scientific, NJ, USA) was dissolved in one mL of 95% ethanol, and diluted to predetermined concentrations with nanopure water (Thermo Fisher Scientific, Waltham, MA, USA). The GA3 solution was applied using a hand sprayer onto the bud and corm surfaces until they were completely wet (about 2.5–3.0 mL for 1- and 2-year-old corms). Distilled water was sprayed in the control treatment. After GA3 application, bulbils and corms were stored at room temperature for one night to facilitate GA3 absorption.
Corms were arranged in a completely randomized complete block design with three replicates. In each replication, 12 corms or 20 bulbils were planted on a planting bed raised 15 cm with a triangle distance of 50 cm × 50 cm × 50 cm. Seed corms were planted 5 cm below the soil surface upside up. At planting, about two grams Furadan 3G® (FMC, Indonesia) was applied around each corm to prevent termite and soil disease infections.
Bud emergence, percentage of flowering plants, time to anthesis, and berry number were evaluated. Morphological characteristics of flower organs were determined at anthesis according to Santosa et al. (2016b). If morphological variations observed in GA3 treatments exceeded the variations found in the control (no GA3 application), they were classified as abnormal variations. Erratic flowering was calculated from the number of inflorescence reversions to vegetative, but vegetative growth after inflorescence death was not considered as erratic. Statistical evaluation was conducted using ANOVA, and for significant data was then analyzed further with Least Significant Difference (LSD) at the 5% level.
Bud, not yet classified either as leaf or flower bud, emergence was significantly affected by interaction among GA3 and corm ages treatments. There was a tendency that larger corms emerged earlier than smaller ones, and bulbil buds emerged later than corms. Fifty percent of bulbil buds emerged 14–16 weeks after planting (WAP) in control, 500 ppm and 1000 ppm, but 50% emergence attained 12–14 WAP with 1500 and 2000 ppm GA3 treatments (Table S1). Buds of 2-year-old corms began to emerge earlier than those of one-year-old corms when corms were treated with 1000–2000 ppm GA3. This finding was in line with the results of previous studies where GA3 could induce flowering in many Amorphophallus species (Itoh, 1960; Latifah and Purwantoro, 2015; Santosa et al., 2006; Zhao et al., 2010, 2013). Vandenbussche et al. (2007) stated that gibberellins can stimulate sprouting, cell expansion and elongation growth. Although the reason for the delay in bud emergence of bulbils in this study is obscure, the absence of main buds in bulbils probably delayed bud emergence.
The percentage of flowering corms increased with increasing corm age and increasing levels of GA3 (Table 1). When GA3 was not applied, the flowering percentage of 2- and 3-year-old corms were 8.3% and 80.6%, respectively, whereas bulbils and 1-year-old corms did not produce any inflorescences. In natural conditions, the flowering percentage of 2- and 3-year-old A. muelleri corms was dependent on corm weight, planting distance, soil moisture, and light intensity (Jansen et al., 1996; Sugiyama and Santosa, 2008; Sumarwoto, 2005; Zhao et al., 2010). According to Sugiyama and Santosa (2008), 10% of 3-year-old plants of A. muelleri flowers under a 25% shading condition, and 45% of flowers under shading of 75%. Zhao et al. (2010) mentioned that more than 48% of 3-year-old corms bore inflorescences naturally.
Percentage of flowering corms, erratic flowering and abnormal flower shape, time to anthesis from planting, and success rate of seed production from different-aged A. muelleri seed corms treated with different level of GA3.
When GA3 was applied, the flowering percentage was affected by corm age (Table 1). For bulbils, 1-, and 2-year-old corms, GA3 application increased the flowering percentage from 1.7–6.7%, 2.8–16.7%, and 11.1–63.9%, respectively, dependent on GA3 levels. This finding supports Jansen et al. (1996) and Zhao et al. (2010) that flower formation is dependent on corm age. According to Zhao et al. (2010), an application of 1 ppm GA3 results in 10% of about 500 g corms (age unknown) and 100% of 200 g corm sections derived from 3-year-old corms that produced inflorescences. In A. paeoniifolius, Sugiyama and Santosa (2008) reported that whole and sections of 3-years-old corms enable flowering naturally in a dark and dry storage room under room temperature. This means that age-dependent flowering in A. muelleri is related to hormonal status, e.g., GA3.
Abnormal inflorescencesApplications of GA3 caused abnormal inflorescences, irrespective of the GA3 levels and corm age (Fig. 1). The incidence rates of abnormal inflorescences were 100%, 50–100%, and 43–67% in bulbils, 1-, and 2-year-old corms, respectively (Table 1).
Inflorescence morphology of A. muelleri from different-aged corms treated with GA3. A, Spathe without limb. B, Co-existence of a leaf and inflorescence from bulbils. C, Co-existence of a leaf and inflorescence from a 2-year-old corm. D, An inflorescence with a pointed spathe apex. E, An inflorescence with a star-shaped spathe. F, Double-spathe, small first spathe and second spathe with multiple apices. G, Evergreen spathe remained until 6 months after anthesis. H, A sheath with a leaflet on the apex. I, An inflorescence bud with elongated sheaths. J, Dull-shaped spathe apex. K, Double-spathe from the control treatment. L, Deformed appendix. M, Deformed reproductive organ. N, Rough appendix. O, petiole showing disturbed geotropism. Bar 1 cm.
Abnormalities were observed in both vegetative (leaf) and reproductive organs (inflorescences) including the elongation of sheaths, sterility and imperfect inflorescences (absence of a male zone and appendix, or spadix) (Fig. 1). All imperfect inflorescences failed to enter anthesis; therefore, the percentage of harvested berries (ratio of berries to female flowers) was low in this type (Table 1). It is unlikely that spathe deformation (star-shaped apex and double spathes; Fig. 1F) typically occurred at specific GA3 levels and corms ages. Morphological abnormalities including evergreen spathe (Fig. 1G), a sheath apex that formed leaflets (Fig. 1H) or geotropism petioles (Fig. 1O) were observed in 1-, and 2-year-old corms applied with GA3.
Morphological abnormalities in inflorescences have previously been reported in several Amorphophallus species (Itoh, 1960; Latifah and Purwantoro, 2015; Santosa et al., 2006; Zhao et al., 2013). Itoh (1960) and Latifah and Purwantoro (2015) found that application of GA3 on cormlets or young corms caused some abnormalities in A. konjac and A. titanum inflorescences, leading to seed production failure. Abnormal inflorescences were observed at a rate of 12.5% and 40% of A. paeoniifolius plants of 2- and 1-year-old corms respectively treated with 2 g·L−1 GA3 (Santosa et al., 2006). In A. muelleri, Zhao et al. (2013) found an abnormality in the stamens when a high concentration of GA3 solution was applied to 2-year-old corms. Abnormalities were also found in Capsicum annuum, and application of GA3 before initiation of floral organs induced abnormalities in petals, feminization of stamens (carpelization) and inhibition of filament elongation (Sawhney, 1981). Moreover, Santosa et al. (2018) compared a population of natural flowering A. muelleri 3-year-old corms with other populations of the same age treated with different levels of GA3 and confirmed that sheath length, spathe disposition and limb apex shape were affected by GA3 applications.
Although GA3 induced flowering in bulbils in the present study, all inflorescences aborted before anthesis, leading to seed production failure (Table 1; Fig. 2). A similar phenomenon occurred in 1-year-old corms unless 1000 and 1500 ppm GA3 solutions were applied. A large number of seeds were produced when GA3 was applied to 2-year-old corms, but GA3 application caused erratic flowering. The erratic flowering was also reported in 3-year-old corms treated with gibberellin, although the number of incidents was low (Santosa et al., 2018). In tropical tree crops like mango (Mangifera indica L.), and lychee (Litchi chinensis Sonn), erratic flowering is common and reduces fruit production (Carr and Menzel, 2014; Yeshitela et al., 2005). In A. muelleri, ability to produce seeds was reduced due to erratic flowering.
Possible effect of GA3 application on different-aged corms of A. muelleri to produce inflorescences (F) and continue life cycle (S). F1 produced a perfect inflorescence. F2 produced some perfect inflorescences and produced seeds (S1), otherwise die (S2). F3 produced mostly imperfect inflorescences that mostly died (S4), and some produced a leaf (S5). S1 success produced seeds, otherwise not. S2 and S3 plants died. S4 and S5 corms continued in a vegetative state and produced new corms. Bar 2 cm, unless otherwise stated.
Singh and Kushwaha (2006) showed that erratic including asynchronous flowering is kind of phenological adaptation to tropical climate variability. Physiologically, Duclos and Björkman (2015) noted gibberellic acid is involved in reproductive transition at the vegetative stage, promoting stem elongation, and inducing reversion toward a vegetative state in cauliflower. In A. muelleri, different phyllotaxis of flower and leaf buds (Sugiyama and Santosa, 2008), may affect failure of some induced corm to maintain a reproductive stage, unlike in cauliflower (Duclos and Björkman, 2015), where bracting around curd is due to growing leaf primordia subtending to the inflorescence meristem. Macroscopic observation of young A. muelleri corm that underwent erratic flowering indicated that if a leaf initially proliferated from the main bud the subsequent flower emerged from the axillary bud, while flowers that initially proliferated from the main bud had subsequent leaves that emerged from the axillary bud (Figs. 2 and 3). It is probable that bud phyllotaxis and sink strength influence sink competition and dominance (Hardwick, 1984; Marcelis et al., 2004), leading to erratic incidents. In addition, A. muelleri leaf is a stronger sink because it is supported by numerous roots, unlike flowering corms (Santosa et al., 2016a, b).
Typical growth stage of a discontinued life cycle (E1–E4) (left) and continued life cycle (G1–G4) (right) of A. muelleri inflorescence after treatment with GA3. A, Leaf emerges followed by spadix. B–C, spadix grows normally. D, Anthesis and berry formation. E1, Peduncle competes for space with leaf and dies. E2, Petiole infected by disease. E3, Petiole dies. E4, Corm rotten. G1, New leaf emerges. G2, Many leaves develop. G3, Leaf reaches maturity. G4, Daughter corm is produced. Arrows show remnants of flowers or infructescence. Bar 5 cm, unless otherwise stated.
Figure 2 summarizes the effect of GA3 applications on corm in relation to plant life cycle. A 3-year-old control corm, mostly produced inflorescences (Fig. 2F1) then produced seeds (Fig. 2S1), but a small number produced leaves (Fig. 2S4). An F2 inflorescence coming from 1- and 2-year-old corms treated with GA3 followed two different paths: (1) spathes were rotten and inflorescences failed to reach anthesis (Fig. 2S2), and (2) an inflorescence continued to develop young berries (Fig. 2S1). An F3 inflorescence coexisting with a leaf shown in Fig. 2 also followed two different paths: (1) if corms failed to produce a new leaf, plants died (Fig. 2S4), (2) corms which developed a leaf produced a new corm at the end of the growing season (Fig. 2S5).
Figure 3 shows an F3 flowering type continuing through the S3 and S4/S5 cycles. In S3, the plant terminated its life cycle due to disease infection at the growing point due to a dead inflorescence (Fig. 3E1–E4). On the other hand, the plant produced new corm in the S4/S5 cycle (Fig. 3G4) after the plant developed new leaves (Fig. 3G1–G3). The mechanism is still unclear regarding how the F3 flowering type chooses S3 or S4/S5 routes, because all these phenomena occurred in both 1- and 2-year-old corms after GA3 application irrespective of the levels.
Flower sizeTime to anthesis was neither affected by corm age nor by GA3 application (Table 1); it ranged from 92.4–102.1 days after planting. The number of emerging sheaths was not significantly different among GA3 levels (2–4 sheaths) at LSD test 5% (Table 2). The peduncle was usually covered by 7–8 sheaths with 4 sheaths remaining underground, whereas the other 3–4 became visible when the inflorescence grew out. Peduncle sheath length was increased with an increase in corm age, irrespective of GA3 concentrations, especially for the last 3 sheaths. In 2-year-old corms, the lengths of the last sheath were 6.5 cm, 8.0 cm, 16.0 cm, 16.1 cm and 16.7 cm in 0, 500, 1000, 1500 and 2000 ppm GA3 treatments, respectively. In 1-year-old corms, the sheath length was about 23–25% shorter than those in 2-year-old corms.
Characteristics of A. muelleri inflorescences from different-aged corms treated with different levels of GA3.
Inflorescence size including peduncle size, spathe length and spadix length at anthesis were affected by corm age, but there were significant interactions between corm age and GA3 levels (Table 2). The largest inflorescences were generally obtained from 2-year-old followed by 1-year-old corms. GA3 application increased peduncle length, but the GA3 concentrations had no effect on peduncle length for 2-year-old corms. Spathe and spadix lengths were significantly longer at the level of 1000 ppm GA3 than 1500 and 2000 ppm GA3 for 2-year-old corms. The ratio of spadix to peduncle length was lower at the level of 1000 ppm GA3 than at the levels of 0 and 2000 ppm, irrespective of corm age (Table 2). One-year-old corms showed a lower ratio of spadix to peduncle length as compared with 2-year old corms. On the other hand, the length of female and male zones (appendix length) was increased with increasing corm age. Appendix length was the longest at the level of 1000 ppm GA3 for 2-year-old corms. Although there was a significant interaction between corm age and GA3 levels for appendix length, there were no significant interactions for the length of female and male zones. This fact indicates some phenotypic plasticity in some A. muelleri inflorescence characteristics is induced by GA3.
Seed productionBerries matured completely 8–10 months from anthesis, irrespective of corm age or GA3 levels. This is largely in line with the results of Santosa et al. (2016b), i.e., 9.6–10.2 months. In this experiment, inflorescences from 1- and 2-year-old corms treated with gibberellin irrespective of the GA3 level exhibited large variations in emerging times (Table S1). Interestingly, some berries of late flowering plants tended to mature at the same time with those of early flowering plants. Consequently, the maturation time of berries from late flowering was shorter. It is likely that maturation relates to the high transpiration rate in the dry season (June–August).
The number of berries per infructescence and number of anthers did not relate to GA3 application (Table 3). The number of berries was significantly higher in old corms than in young corms, irrespective of GA3 treatments; it ranged from 48–90, and 96–139 for 1- and 2-year-old corms, respectively. The number of anthers from 1-year-old corms was 200–586, and from 2-year-old corms from 819–1005. In controls, 3-year-old corms had 1530–1735 anthers. The present findings are unlike in other species such as the parthenocarpic tomato, where blocking gibberellin biosynthesis using paclobutrazol promotes seed production (Ohkawa et al., 2012); in Cucurbitaceae such as Cucumis sativus L and Momordica charantia L, application of GA3 at 400 ppm stimulates the number of staminate and pistillate flowers, i.e., 37–59% and 43–44%, respectively (Khan and Chaudhry, 2006). Cheng et al. (2013) noted that GA3 application in grapes causes embryo abortion and stimulates high seedless rates. The effectiveness of GA3 to alter seed production in apomicts species like A. muelleri requires further study.
Number of berries, seeds and anthers in A. muelleri inflorescences from different-aged corms treated with different levels of GA3.
This study showed that the inflorescence size and seed production were dependent on mother corm size or corm age. It seems that large inflorescences produce a larger number of seeds than small ones. Because A. muelleri corm contains a large amount of glucomannan (Jansen et al., 1996), the dependency of inflorescence size on corm size suggested that the activity of glucomannan-degrading enzymes such as β-mannanase (Chua et al., 2013) could be different with corm size or corm age. It is known that sugar concentration is important for successful flower development (Christiaens et al., 2016; Matsoukas et al., 2013; Tooke et al., 2005). According to Sumarwoto (2005), the glucomannan content of bulbils of 1-, 2- and 3-year-old corms ranged from 25–30%, 35–39%, 46–48%, and 47–55% of dry mass, respectively, and the glucomannan decreased by 15–20% during flower development. It is probable that flowering abnormalities including lack of ability to maintain the reproductive phase in induced-young corms is due to a lack of starch turnover. Matsoukas et al. (2013) stated that lack of starch turnover causes low carbohydrate availability.
According to Chua et al. (2013), glucomannan could be decomposed in the corm of A. konjac by the action of β-mannanase. Previously, Shimahara et al. (1975) identified two β-mannanases, I and II, from sprouting A. konjac corms and revealed that both β-mannanase I and II could hydrolyze β-1,4-mannooligosaccharides and glucomannooligosaccharides. In germinated tomato seeds, Nanogaki et al. (2000) identified two genes responsible for β-mannanase activity as LeMAN1 and LeMAN2, for which activity was affected by gibberellin, whereas Toorop et al. (1998) reported that β-mannanase activity was low at high abscisic acid (ABA) concentrations. Gupta and Kaur (2000) stated that catabolism in seeds with a high mannan content is affected by gibberellic acid. Although it is not clear whether corm size or corm age affect the endogenous gibberellic acid and ABA contents and the responsiveness to gibberellic acids, GA3 could induce high β-mannanase activity in 2-year-old corms, resulting in larger inflorescences than 0- and 1-year-old corms.
The response of A. muelleri to GA3 applications is different from A. paeoniifolius. In A. paeoniifolius, Santosa et al. (2006) reported all small cormlets produced flowers after GA3 application, irrespective of the GA3 levels. On the other hand, flowering induction at non-flowering age in A. muelleri was unpredictable after GA3 application, indicating corm size is related to flower production. Erratic flowering became apparent in younger corms (bulbils and 1-year-old corms). In Satsuma mandarin, erratic floral production in Japan is related to expression of FLOWERING LOCUS T expression (Nishikawa et al., 2017). In the present experiment, two types of erratic flowering were observed, 1) initially corm produced a leaf followed by an inflorescence, and 2) initially produced an inflorescence followed by a leaf(s). The second type refers to the reversion after Tooke et al. (2005), i.e., inflorescence reversion is due to genetic, environmental and physiology factors such as partial induction, meristem plasticity, perennial type and pseudovivipary. In Arabidobsis suecica, Asbe et al. (2015) concluded flower reversion to a vegetative state is related to day length, timing of flowering and geographic origin. In longan fruit trees, Jia et al. (2014) stated that continuous flowering may involve SVP, GI, FKF1 and ELF4 gene expression.
Erratic flowering clearly reduces the success rate of seed production in enforced flowerings. However, finding on inflorescence reversion phenomenon in A. muelleri due to GA3 application need further study. In corm fields, farmers emasculate flowers in order to maintain high corm production (Santosa et al., 2003). Zhao et al. (2010) showed that flowering corms contain 25–34% lower glucomannan levels than 3-year-old corms without flowers. Thus, the possibility to reverse flowering in 3-year-old corms is a future challenge in A. muelleri in order to improve field corm production.
ConclusionThe effectiveness GA3 application to induce flowering in A. muelleri depended on corm age, and increasing GA3 levels increased the percentage of flowering, irrespective of corm age. However, GA3 applications at levels of 500 and 2000 ppm caused 100% abnormal flowering in bulbils and 1-year-old corms due to vigorous leaf growth prior to, and after, flowering, resulting in low levels of seed formation. We also found that the flowering time that usually occurs after three years of planting could be reduced, although some erratic flowering occurred in 2-year-old corms. This implies that seeds could be produced by the application of 1500 ppm GA3 to 2-year-old corms of A. muelleri.
This is a collaborative research project between Bogor Agricultural University, Indonesia and Tokyo University of Agriculture, Japan.
