Introduction

Phalaneopsis is one of the most valuable potted ornamental plants produced around the world. However, in some orchid nurseries, floral bud drop has become a big problem affecting quality and commercial value of the plants. Numerous studies have reported floral bud drop in popular tropical orchids grown for cut flowers, and consequently, many farmers in South-East Asia have stopped growing Aranthera Beatnice Eng and Dendrobium Sri Siam because of this bud drop problem (Hew and Clifford 1993). Less information is available on bud drop in miniature orchids.

It is well known that floral buds of various orchids are sensitive to ethylene (Hew and Clifford 1993). In Dendrobium, it has been observed that abscission of floral buds and flowers took place frequently during transportation in sealed cardboard boxes. This induced the accumulation of ethylene produced from plants inside the box resulting in abscission of buds and flowers. Most of the ethylene originated from floral buds rather than from open flowers (Uthaichay et al. 2007). Pretreatment of Dendrobium with 1-MCP (1-methylcyclopropene), an ethylene inhibitor, had a strong effect of preventing the abscission of floral buds and open flowers. A similar effect was observed in plants treated with STS (silver thiosulphate). 1-MCP treatment of Dendrobium not only inhibited ethylene action, but also limited the ethylene production by limiting ACC (1-aminocyclopropane-1-carboxylic acid) synthase activity and rate of ACC conversion to ethylene (Uthaichay et al. 2007). In Cymbidium, ethylene caused purple discoloration of the rostellum and induced flower senescence. 1-MCP extended the vase life of Cymbidium flowers, both exposed to air and ethylene (Heyes and Johnston 1998). Similar results were also obtained in a study with Cattleya (Yamane et al. 2004).

Like in many other flowers, the phenomenon of pollination-induced flower senescence also happens in orchid flowers. After pollination, cut Phalanenopsis flowers undergo rapid acceleration of the wilting process after 24 h. There was an increase in ethylene sensitivity occurring 4 h after pollination, reaching a peak 10 h after pollination, and all these changes happened before any increase in ethylene production (Porat et al. 1994). Additionally, it has been suggested that pollination-induced changes in ethylene sensitivity in Phalaenopsis flowers are the cause triggering autocatalytic production of ethylene to its climacteric peak (Porat et al. 1995). This sensitivity increase can be limited by treating flowers with the EDTA (ethylenediaminetetraacetic acid), a calcium ion chelator, and the increase of ethylene production in flowers was thus delayed efficiently (Porat et al. 1995).

Based on the findings above, ethylene could be the potential reason for and play a role in regulating bud drop processes in orchids, but presently the causes of floral bud drop still remain unknown. In numerous ornamental flowers such as roses, Begonia and Hibiscus, endogenous and exogenous ethylene are the causes of bud and flower abscission (Høyer 1984, 1985; Müller et al. 1998; Woltering and van Doorn 1988).

In the present study, the performance of different Phalaenopsis cultivars and their physiological responses after exposure to ethylene was investigated. It was hypothesized that water loss, membrane permeability and the increase of ABA concentration in petals are contributing to bud and flower drop. Additionally, it was anticipated that different cultivars have different sensitivity to ethylene and that treatment with the ethylene blocker 1-MCP could inhibit the ethylene-induced abscission of floral buds.

Materials and methods

Plant material

The mini Phalaenopsis cultivars (in 6 cm pot) Sogo ‘Berry’, Sogo ‘Gotris’, Sogo ‘Vivien’ and Sogo ‘Yenlin’ used in this experiment are breeding lines from Sogo team, Taiwan. They were grown in a Danish nursery under greenhouse condition at 20–21°C, RH 60–70% and a light intensity 125–175 μmol m−2 s−1. At the stage of the commercial maturity (with maximal 1–2 open flowers), the plants were transported to the university within 24 h. They were placed immediately in growth chambers with controlled climate simulating interior conditions at 21°C and 12 h light from cool white fluorescent lamps (50 W, Philips, Holland) providing 15 μmol m−2 s−1 PAR (photosynthetic active radiation).

Exposure to exogenous ethylene

Two independent experiments with different purposes were carried out. The first experiment was conducted with the Phalaenopsis cultivars Sogo ‘Berry’, Sogo ‘Gotris’, Sogo ‘Vivien’ and Sogo ‘Yenlin’ for investigating the ethylene sensitivity of selected cultivars. The plants were placed in three identical glass chambers and treated with exogenous ethylene at the concentrations 0, 0.1, 0.5 ppm, respectively. The 124 l glass chambers were tightly sealed so that no leakage of gas occurred during the treatment. The glass chambers were located in growth chambers with controlled climate simulating interior conditions, i.e. a temperature of 21°C, photoperiod of 12 h light from cool white fluorescent lamps (50 W, Philips, Holland) with a PAR providing 15 μmol m−2 s−1. Four plants of each cultivar were placed randomly inside the glass chamber. The chambers were opened and ventilated every day, then sealed again and the same concentration of ethylene was injected for each treatment. The effect of ethylene on display quality was determined by daily observation of color, turgidity and abscission of the floral buds and flowers. After 8 days of ethylene exposure, the number of healthy flowers/buds and bud/flower drop were recorded individually. Subsequently, the percentage of healthy flowers and buds and the rate of bud and flower drop were calculated. The experiment was replicated twice.

The second experiment was conducted for investigating physiological changes in plants exposed to exogenous ethylene and their relationship to floral bud drop. The same cultivars as used in the first experiment were treated with 0 and 0.5 ppm ethylene. All other conditions during the treatment were identical to the first experiment. Buds and flowers were excised for analysis every second day and at the end of the experiment. Bud/flower water content and electrolyte leakage were determined at each sampling. Additionally, ABA content in petals was measured in samples collected at the end of experiment.

1-MCP pretreatment

The Phalaenopsis cultivar Sogo ‘Yenlin’ was chosen to investigate the effect of the ethylene inhibitor 1-MCP on plants, because of orchid industries’ interest in solving this cultivar’s bud drop problem. Ten plants were sealed in a 124 l glass chamber at 21°C and 12 h light from fluorescent tubes providing 15 μmol m−2 s−1 PAR. 1-MCP was released from a commercial powdered formulation as described by the producer (SmartFresh powder, 0.14%, AgroFresh, USA) and the amount of powder required to obtain a concentration of 500 ppt was placed in a beaker and 0.6 ml water was added to release gaseous 1-MCP. The glass chamber with plants was immediately sealed after the water was added. The plants were exposed to the inhibitor overnight and ethylene treatment started on the next day. Half of the plants were exposed to 0.5 ppm ethylene in a sealed chamber, together with the same number of plants without 1-MCP pretreatment. The other half of the 1-MCP pretreated plants were sealed in a chamber as well but exposed to air together with the same number of plants without 1-MCP pretreatment. During the ethylene treatment, the chambers were opened every day, ventilated and supplied with the same concentration of ethylene. The performance of plants was evaluated after 5 days ethylene or air exposure. The number of healthy flowers, healthy buds and bud and flower drop were recorded individually. The experiment was replicated twice. Water content, electrolyte leakage and ABA content of petals were measured at the end of the observation period.

Water content in petals

Two flowers or buds were excised from each plant and immediately transported to the laboratory. After fresh weight determination, the flowers or buds were oven dried to a constant weight at 80 °C, and water content was calculated on fresh weight base.

Electrolyte leakage in petals

Two flowers or buds were excised from each plant and weighed immediately. Petals were immersed into beakers containing 25 ml miliQ water. The beakers were placed on the laboratory shaker for 1 h at room temperature. Electrolyte leakage was determined by the changes in conductivity using a conductivity meter (YSI 3200, Yellow Spring, USA). The conductivity was calculated on the fresh weight base.

ABA content in petals

ABA content in petals was determined in both experiments of exposure to exogenous ethylene in four cultivars and in the experiment of 1-MCP pretreatment. Two to three buds or flowers were excised from each plant and were immediately stored in liquid nitrogen. Plant material was ground with liquid nitrogen and after being weighed transferred into a 15 ml centrifuge tubes. Four ml phosphate buffer (0.02 M PO4 3−, pH 7.3) was added to each tube and boiled for 5 min in a water bath. The samples were shaken overnight at 4°C on a lab shaker (VXR basic IKA Vibrax, IKA, Germany) with a speed of 1,200/min. The extracts were centrifuged and 1 ml of the supernatant was transferred into another tube. One ml freshly prepared water-saturated ethyl acetate was added into the tube followed by adding 3 drops of 1 M H2SO4. After phase separation in each sample tube, the organic phase of each sample was transferred to silica columns (Sep-Pak Vac 3CC, Water Corporation, USA). This procedure was repeated twice to make sure all the extracted ABA was transferred. The samples were air dried and dissolved in 1 ml miliQ water for ABA determination.

ABA content in the samples was determined by ELISA (Enzyme Linked Immuno Sorbent Assay) following the protocol described by Asch (2000). In order to examine if the extracted samples contain compounds which react with the antibody other than antigen (ABA), a cross-reaction test was done before conducing the real measurement (Quarrie et al. 1988). It was shown that there was no cross-reaction in the sample extract.

Experimental design and statistics

For the study of ethylene sensitivity in four selected Phalaenopsis cultivars in 0, 0.1, 0.5 ppm ethylene, two replications were conducted and four plants were used for each experiment. To measure water content, electrolyte leakage and ABA content in petals in four cultivars in air and 0.5 ppm ethylene, two replications were conducted with four plants of each cultivar per treatment. In the experiment of investigating the effect of 1-MCP on the cultivar Sogo ‘Yenlin’, two replications were conducted with five plants per treatment. Data were subjected to analysis of variance (ANOVA) procedures (SAS Institute, Inc. 1988). Standard error of the means (SE) was calculated. Differences among means were determined using independent t-test.

Results

Performance of Phalaenopsis cultivars after exposure to exogenous ethylene

Exogenous application of ethylene to mini Phalaenopsis resulted in a significant decrease in plant quality in all cultivars. Sogo ‘Vivien’ was most sensitive to ethylene treatment and showed symptoms of floral bud and flower senescence in response to 0.1 ppm ethylene after only 3 days of treatment.

During and after 5 days air exposure, none of the cultivars showed any symptoms of decreasing quality. In response to ethylene, firstly all investigated cultivars showed symptoms of color change on either buds or flowers, then water loss took place and petal tissue appeared wrinkled, and finally plants reacted with bud or flower drop (Fig. 1). Plants exposed to the lowest ethylene concentrations of 0.1 ppm exhibited clear symptoms but the treatment resulted in distinct differences in the percentage of healthy flowers and buds among the four cultivars. Cultivar Sogo ‘Vivien’ was extremely sensitive to ethylene and all flowers and buds were affected. In this sensitive cultivar quality was lost both in response to 0.1 ppm and 0.5 ppm ethylene. Compared to the other cultivars, Sogo ‘Berry’ was relatively tolerant to ethylene and exhibited a higher percentage of healthy flowers and buds of 41% at 0.1 ppm and 12.5% at 0.5 ppm. All investigated cultivars exhibited symptoms of bud drop at the end of exposure to ethylene. Sogo ‘Vivien’ showed 100% bud abscission even at 0.1 ppm ethylene. On the other hand, Sogo ‘Berry’ had about 40% bud drop at 0.1 ppm and 60% bud drop at 0.5 ppm ethylene while the rest of the cultivars lost all of their buds (Table 1; Fig. 2).

Fig. 1
figure 1

Symptoms of the cultivar ‘Yenlin’ exposed to 0.5 ppm ethylene (left) compared to air control (right) after 8 days exposure. Ethylene affected buds appeared reddish, wrinkled and were losing gloss surface compared to unaffected buds

Full size image
Table 1 Differences in the percentage of healthy flowers and buds in four commercial cultivars of mini Phalaenopsis Sogo ‘Berry’, Sogo ‘Gotris’, Sogo ‘Vivien’ and Sogo ‘Yenlin’ exposed to a range of concentrations of ethylene for 8 days
Full size table
Fig. 2
figure 2

Buds and flower drop in four cultivars of mini Phalaenopsis Sogo ‘Berry’, Sogo ‘Gotris’, Sogo ‘Vivien’ and Sogo ‘Yenlin’ after exposure to a range of concentrations of ethylene after 8 days. Values are mean of 4 replicates. Error bars represent ± SE

Full size image

Changes of water content, electrolyte leakage and ABA content in petals

After 8 days of exposure to ethylene or air, respectively, all cultivars showed significant water loss in ethylene but none of them in air. Among the four cultivars, ‘Gotris’ had significantly lower water content compared to ‘Berry’ and ‘Yenlin’ and lost 12% of water in petals after 8 days exposure to 0.5 ppm ethylene. No significant difference in water loss was found between ‘Gotris’ and ‘Vivien’. The cultivar ‘Berry’ had the significantly highest water content after ethylene treatment (Fig. 3).

Fig. 3
figure 3

Changes of water content in petals of four cultivars of mini Phalaenopsis Sogo ‘Berry’, Sogo ‘Gotris’, Sogo ‘Vivien’ and Sogo ‘Yenlin’ during 8 days exposure to 0.5 ppm ethylene or air. Values are mean of 4 replicates. Error bars represent ± SE. Different letters above bars indicate significant difference (P < 0.05) in two treatments (exposure to 0.5 ppm ethylene or air) according to independent t-test

Full size image

None of cultivars showed a significant increase of electrolyte leakage from petals in air. In contrast, ethylene treatment resulted in a higher rate of electrolyte leakage from petals after 4 days and the measured conductivity reached to 55–210 μs/cm g after 4 days compared to 13–20 μs/cm g in the beginning, depending on different cultivars. The cultivar ‘Vivien’ showed the biggest damage of membranes while ‘Berry’ was least damaged (Fig. 4).

Fig. 4
figure 4

Changes of electrolyte leakage in petals of four cultivars of mini Phalaenopsis Sogo ‘Berry’, Sogo ‘Gotris’, Sogo ‘Vivien’ and Sogo ‘Yenlin’ during 8 days exposure to 0.5 ppm ethylene or air. Values are mean of 4 replicates. Error bars represent ± SE. Different letters above bars indicate significant difference (P < 0.05) in two treatments (expose to 0.5 ppm ethylene or air) according to independent t-test

Full size image

There was a close relationship between water content and electrolyte leakage in petals (Fig. 5). Along with the decrease of water content, the electrolyte leakage was increasing and the relationship was highly significant (P < 0.01).

Fig. 5
figure 5

Relationship between water content and conductivity of floral petals after expose to 0 and 0.5 ppm ethylene in four cultivars of mini Phalaenopsis Sogo ‘Berry’, Sogo ‘Gotris’, Sogo ‘Vivien’ and Sogo ‘Yenlin’

Full size image
Fig. 6
figure 6

ABA content in petals of the four mini Phalaenopsis cultivars Sogo ‘Berry’, Sogo ‘Gotris’, Sogo ‘Vivien’ and Sogo ‘Yenlin’ after 8 days exposure to 0.5 ppm ethylene or air. Values are mean of 4 replicates. Error bars represent ± SE. Different letters above bars indicate significant difference (P < 0.05) in two treatments (exposed to 0.5 ppm ethylene or air) according to independent t-test

Full size image

Ethylene-treated petals contained a higher ABA content than those exposed to air. However, only the cultivar ‘Gotris’ showed a significant difference between the two treatments (air and ethylene). In the other cultivars, the observed higher ABA content in response to ethylene exposure was not significant. Concerning the comparison among cultivars, ‘Gotris’ contained a significantly higher ABA content than the other cultivars, both in air and ethylene.

The effect of 1-MCP

Mini Phalaenopsis Sogo ‘Yenlin’ pretreated with 500 ppt 1-MCP exhibited resistance to the deleterious effects of ethylene. While plants without 1-MCP pretreatment lost more than 80% flower quality after exposure to 0.5 ppm ethylene, no quality loss took place for the plants exposed to ethylene after 1-MCP pretreatment and the ones in air (Fig. 7).

Fig. 7
figure 7

Performance of Sogo ‘Yenlin’ in different treatments from left to right: air, 0.5  ethylene, 500 ppt 1-MCP + 0.5 ppm ethylene, and 500 ppt 1-MCP + air

Full size image

With respect to water content in petals, the ethylene-treated plants had the lowest water content compared to other treatments (Fig. 8). Furthermore, the water content in petals exposed to air treated by 1-MCP was significantly higher than the ones exposed to air but without 1-MCP pretreatment.

Fig. 8
figure 8

Water content (a), electrolyte leakage (b) and ABA content (c) in petals of mini Phalaenopsis Sogo ‘Yenlin’ in response to different treatments after 8 days exposure: 0.5 ppm ethylene, 500 ppt 1-MCP + 0.5 ppm ethylene, air, 500 ppt 1-MCP + air. Values are mean of 5 replicates. Error bars represent ± SE. Different letters above bars indicate significant difference (P < 0.05) among different treatments according to independent t-test

Full size image

In regard to the membrane integrity, ethylene treated plants without 1-MCP pretreatment showed greatest damage in their petal membranes and the measured electrolyte leakage was about 10 times higher than the other treatments (Fig. 8).

The experiment with 1-MCP pretreatment also showed that ethylene treated plants without 1-MCP pretreatment had a significantly higher ABA content than the rest of treatments. There was no significant difference of ABA content among the three treatments: ethylene with 1-MCP, air and air with1-MCP (Fig. 8).

Discussion

During recent years production of Phalaenopsis has been going up significantly, and especially cultivars applicable as mini Phalaenopsis have found growing interest (Bourgeois 2008). However, floral bud drop problems of mini Phalaenopsis plants during production can give rise to significant quality losses and thereby reduce the commercial value of the production considerably. Additionally, bud and flower drop also can reduce quality during postharvest handling and thereby making marketing complicated and reduce pricing.

The aim of the present investigation was to enlarge existing knowledge related to bud drop problems in Phalaenopsis, especially to study the role of ethylene in the problem and to investigate if the ethylene action inhibitor 1-MCP could be used to avoid ethylene-induced quality losses. In our investigation, mini Phalaenopsis was shown to be sensitive to ethylene and we identified bud drop as the most crucial symptom. This observation indicates the possibility that both during production and postharvest periods, factors related to ethylene could be the potential reason inducing bud drop.

Other ethylene-induced symptoms were color changes and water loss in buds and flowers with subsequent loss of gloss, appearance of wrinkles and wilting. The symptoms of color changing and wilting after exposure to ethylene were similar as described in an earlier study in the family Orchidaceae (Woltering and van Doorn 1988). In this fundamental study, orchids were characterized as flowers that respond to ethylene by hastening senescence of petals that initially stay attached to the flower (Woltering and van Doorn 1988). This phenomenon was also observed in our study and no petal abscission was observed but whole floral buds or flowers abscised. On the other hand, the wilting process and significant water loss as response to exogenous ethylene in our study are analogous to pollination-induced senescence of Phalaenopsis (Porat et al. 1994). In contrast to roses and Kalanchoe, where big differences in sensitivity among cultivars were found (Müller et al. 1999; Serek and Reid 2000), the differences in ethylene sensitivity among cultivars in the present study appear less dramatic. However, our study included only four cultivars, and future studies comprising more cultivars may give a more detailed picture of the variability in ethylene sensitivity among cultivars.

Loss of membrane integrity was observed in all Phalaenopsis cultivars exposed to exogenous ethylene and there were distinct differences among cultivars (Fig. 4). This may suggest that the observed bud drop is the result of damage and senescence of petal membranes induced by ethylene, in accordance with early findings that ethylene can enhance the membrane senescence in flowers (Marangoni et al. 1996) and in Tradescantia, where the application of ethylene hastened the onset of an increase of membrane permeability (Suttle and Kende 1980). Furthermore, in cut carnations the reduction of water content in petals appeared to be an early senescence symptom associated with increased electrolyte leakage in petals, which led to great damage of petal membranes (Eze et al. 1986). This finding has been partly confirmed in our study in Phalaenopsis and the close relationship between water content and electrolyte leakage is shown in Fig. 5. There are several noteworthy factors related to ethylene sensitivity in the studied cultivars. For example, the most ethylene sensitive cultivar ‘Vivien’ showed a significantly higher electrolyte leakage after exposure to ethylene than the other cultivars and also relatively higher water loss than the other cultivars except ‘Gotris’. However, there is no statistical difference of water loss between ‘Gotris’ and ‘Vivien’. On the other hand, the relatively insensitive cultivar ‘Berry’ had lowest water loss and electrolyte leakage in response to ethylene (Figs. 3, 4).

Furthermore, the increase of ABA content in petals was one of our most intriguing observations studying the effect of ethylene on Phalaenopsis. In our investigation, all cultivars exhibited a weak increase of ABA content after exposure to ethylene, and in the cultivar ‘Gotris’ the difference in ABA content between ethylene treated and control was significant (Fig. 6). It can be speculated that the fact that petal samples were harvested from flowers in different stages of floral development could have made the difference less clear in the other cultivars. A difference in ABA production depending on flower stage has been documented in roses, where the cultivar ‘Bronze’ had a higher ABA content in floral buds but lower ABA content in newly opened flowers. After flower opening, ABA content in flowers increased during senescence (Müller et al. 1999). A similar phenomenon was also found in the study of carnation (Eze et al. 1986). However, even though not all cultivars exhibited a significant ethylene-induced increase in ABA content, the experiment with 1-MCP pretreatment in the cultivar ‘Yenlin’ also indicated that ethylene treatment can raise ABA level in flowers (Fig. 8). On the other hand, the difference in ethylene-induced ABA content can also be due to cultivar differences in ABA levels. In the rose cultivar ‘Bronze’ a different ABA content at different developmental stages of the flowers were found, while the cultivar ‘Vanila’ had a constant low ABA content at all floral stages (Müller et al. 1999).

There are not many studies in orchids dealing with the interaction of ABA and ethylene. Application of ABA to Cymbidium flowers induced some, but not all, post-pollination symptoms such as turgidity loss (Arditti et al. 1970). Wang et al. (2002) reported that the decrease of ABA content in roots and buds of Phalaenopsis hybrida was correlated with bud activation and the development of flowering shoots. Dormant floral buds and roots in Phalaenopsis hybrida contained a relatively higher level of free ABA and free ABA levels in the roots and buds decreased until flowering shoot initiation. Applying exogenous ABA significantly inhibited spiking, and this inhibition was more pronounced with increasing ABA doses. In Dendrobium during the process of low temperature-induced flower initiation, the amount of ABA decreased in both leaves and buds (Campos and Kerbauy 2004). However, none of the studies has been related to the senescence of Phalaenopsis and the interaction with ethylene.

In an earlier study Mayak and Halevy (1972), assumed that ethylene might have induced or at least mediated the synthesis of ABA in petals of roses. Müller et al. (1999) documented that ethylene can elevate ABA content in roses and found differences in ABA content in cultivars with different flower longevities. In another study in roses, ABA and ethylene both stimulated senescence and were suggested to interact with flower senescence under water stress (Kumar et al. 2008). In the present investigation, we suggest that the increase of ABA content could be a response to water loss in petals, which was also assumed by Eze et al. (1986) in a study on carnation flowers. In our study ‘Gotris’ had significantly higher water loss than the other cultivars, except ‘Vivien’, and at the same time ‘Gotris’ exhibited a significantly increased ABA content. However, it remains unclear if the increase in water loss induced ABA content, or vice versa the higher ABA content was the reason for a higher water loss in petals and increased electrolyte leakage. ABA might increase the sensitivity of flowers to ethylene and intensify the senescence process, which was also suggested by Ronen and Mayak (1981) in their study in carnations. However, there was no clear correlation between ethylene sensitivity and ABA found in our study and different cultivars contained different levels of ABA (Fig. 6). The degree of ethylene-induced changes in ABA content and vice versa the impact of ABA content on ethylene sensitivity and thereby promotion of senescence may be cultivar dependent. Here we show that exogenous ethylene increases bud ABA content and promotes senescence and flower abscission of Phalaenopsis orchids. Further studies are required to elucidate how the hormones ABA and ethylene act in concert to regulate water status and membrane permeability during flower senescence.

The ethylene inhibitor 1-MCP was shown to be an important practical tool in inhibiting ethylene action as hypothesized. Due to the interest from industry, the cultivar ‘Yenlin’ was chosen to investigate the effect of 1-MCP. The present results clearly demonstrated that 1-MCP can prevent ethylene action and inhibit senescence processes such as reduced water content, increased membrane permeability and elevated ABA content. Inhibiting effects of 1-MCP on ethylene was also observed in studies with Dendrobium (Uthaichay et al. 2007) and pollinated-induced ethylene production in Phalaenopsis (Porat et al. 1995), but none of them mentioned the inhibiting effects on membrane permeability and ABA content. In the present investigation of the effect of 1-MCP, the ethylene-induced increase in ABA content was prevented in petals of the cultivar ‘Yenlin’ (Fig. 8). This result indicates the close relationship between ethylene and ABA during senescence of Phalaenopsis. Similar effects have been observed in the study of roses where 1-MCP delayed ABA-induced flower senescence (Müller et al. 1999). The effect might be indirectly through inhibiting the decrease of water loss in petals, but more research is required to elucidate the underlying mechanisms.