2020 Volume 89 Issue 4 Pages 367-374
In fruit production, fruit thinning is required to maximize fruit quality and to protect the mother trees. However, thinning is troublesome and laborious work. Fruit self-thinning is the spontaneous elimination of pollinated flowers or fruits within a week to a month after pollination. Since the fruit self-thinning trait has the potential to improve fruit trees, a number of studies using fruit crops such as apple, orange, and tomato have been conducted to clarify the underlying mechanisms. The Japanese pear accession ‘Chojuro’ and its descendants ‘Niitaka’, 162-29, and ‘Akiakari’ have this trait. To identify the pattern and trigger of thinning in the accessions, we pollinated all flowers on the flowering day and recorded the numbers of retained and abscised fruits and the order of flowering in the cluster. The number of retained flowers/fruits in a cluster was widely variable in ‘Chojuro’ and ‘Niitaka’, but was uniform at 3 to 6 fruits per cluster in 162-29 and ‘Akiakari’. In 162-29 and ‘Akiakari’, the earlier the flower opened, the more likely it was to be retained in the cluster, similar to previous observations in apple. In contrast, ‘Chojuro’ and ‘Niitaka’ fruits abscised independently of the flowering order. Therefore, the pattern of fruit self-thinning in pear depends on the accession. To identify the trigger for fruit self-thinning, we analyzed changes in the levels of endogenous auxins and abscisic acid. The results implicate auxin and, to a lesser extent, abscisic acid in fruit self-thinning. Retained fruits showed temperature-dependent transient auxin accumulation, which may trigger self-thinning in pear.
Fruit thinning is widely practiced in many crops, including apple (Malus × domestica Borkh.), pear (Pyrus spp.), peach (Prunus persica (L.) Batsch), and orange (Citrus sinensis) in order to maximize fruit quality and size and to avoid biennial bearing (Looney, 1993; Dennis Jr., 2000; Link, 2000). Hand thinning is the safest method because other mechanical methods risk over-thinning or damaging fruits. However, hand thinning is so labor- and cost-intensive in peach that it is unlikely to become standard orchard management practice (Costa and Vizzotto, 2000).
Fruit self-thinning is the spontaneous elimination of young fruits within a week to a month after pollination. In apple, terminal flowers that become king fruits open and set fruits first in the cluster, although most of the lateral fruits are abscised (Bangerth, 2000). In orange, terminal flowers also open first, but they do not dominate over lateral fruits, and the pattern of abscission is complex (Spiegel-Roy and Goldschmidt, 1996). These observations suggest that the pattern of fruit self-thinning is not uniform and depends on the species.
Several factors are considered candidates for stimulating young fruit abscission. These include a source–sink relationship, endogenous levels and flows of plant hormones, and night temperature (Bangerth, 2000; Costa et al., 2018). A number of studies in various fruit crops have reported the effects of plant hormones as thinning agents (Costa and Vizzotto, 2000; Guardiola and García-Luis, 2000; Wertheim, 2000). Among then, auxins are strong candidates as triggers for fruit self-thinning. In apple, the following mechanism of fruit self-thinning has been suggested: auxin exported from young fruit increases shortly after fruit set, and auxin transported to the abscission zone inhibits auxin transport from lateral fruits, resulting in their abscission (Bangerth, 1989; Gruber and Bangerth, 1990). Abscisic acid (ABA) is another plant hormone thought to regulate abscission. A recent study suggested that ABA induces fruit abscission primarily indirectly by reducing carbon assimilation via stomatal closure (Einhorn and Arrington, 2018). Ambient temperature may also be a key to stimulate abscission. In apple, the abscission rate is increased by high nighttime temperature, although the strength of the effect varies among cultivars (Bangerth, 2000).
In Japanese pear (Pyrus pyrifolia Nakai), the accessions ‘Chojuro’, ‘Niitaka’, 162-29, and ‘Akiakari’, but not ‘Kosui’, have the fruit self-thinning trait (Saito et al., 1998). ‘Niitaka’, 162-29, and ‘Akiakari’ are descendants, in this order, of ‘Chojuro’ (Fig. 1). In apple, phenotypic and genotypic correlation analysis using F1 progeny from a cross between a hybrid and a cultivar with different ideotypes suggested that the mean number of fruits per inflorescence and the percentage of inflorescences with one fruit are influenced by the genotype (Celton et al., 2014). In pear, the mechanism of fruit self-thinning is largely unknown. In this study, to obtain more information on fruit self-thinning in Japanese pear, we observed its patterns. We focused on the number and position of retained fruits within a cluster, which are important for commercial cultivation, and investigated the levels of endogenous auxin and ABA before self-thinning to identify a trigger.
Pedigrees of Japanese pear accessions used in this study. ‘Chojuro’ and its self-thinning descendants are indicated in solid boxes. PI (Plant Introduction) numbers are noted in parentheses.
Five kindred Japanese pears (Pyrus pyrifolia Nakai); ‘Chojuro’ (Plant Introduction number PI297347), ‘Niitaka’ (PI392317; offspring of ‘Chojuro’ × ‘Amanogawa’), 162-29 (‘Niitaka’ × ‘Hosui’), ‘Akiakari’ (162-29 × 42-6), and ‘Kosui’ (PI235634) were used (Fig. 1). 162-29 was raised from seeds, and the other cultivars were grafted onto a rootstock. The trees were grown and maintained at the orchard of the Institute of Fruit Tree and Tea Science, NARO (Ibaraki, Japan) and were about 20 years old with 25 to 30 m2 per tree at the time of this study. Trees were trained on horizontal trellises used in commercial production in Japan and managed in accordance with standard orchard practices.
Investigation of flowering date and fruit setWe used 19 to 29 clusters and 136 to 206 flowers per accession (Table 1). The flowers of each cultivar were pollinated by hand with mixed pear pollen grains obtained from several accessions on the flowering day to ensure that all flowers were pollinated. The pollinated flowers were labeled with colored threads to indicate the flowering date, and each cluster was labeled with colored tapes to identify the flowering date of the first flower to open within it (Fig. S1a). Flowers were pollinated every day until all had opened. The number and flowering date of flowers within clusters were recorded, and their flowering order was noted. A month later, the number and position of the fruits retained were recorded. ‘Chojuro’ and 162-29 were analyzed in 2008 and 2009 and the other four accessions in 2009.
Fruit self-thinning in six Japanese pear accessions in 2008 and 2009. Clusters and flowers were counted on flowering day, and retained fruits were counted 1 month later.
Flowers, fruits, or spurs 6, 8, 14, and 21 days after pollination (DAP) in 2008 and 4, 6, 8, 11, 14, and 21 DAP in 2009 were collected. Flowers and fruits were collected minus stamens, petals, calyx, and peduncle (Fig. S1b). Spurs were sampled from 162-29 in 2008 by cutting them from the stem and stripping off all flowers and fruits, including pedicles (Fig. S1c). At least 10 flowers/fruits or spurs were mixed in one sample. In 2008, two treatments were used for 162-29: (1) unthinned clusters were left intact to permit self-thinning, and (2) thinned clusters were hand-thinned to two flowers per cluster on the day of first flowering to avoid self-thinning.
For indole-3-acetic acid (IAA) quantification, samples were extracted with a mortar and pestle in 3 mL of 80% methanol containing 0.1 mg·mL−1 2,6-di-tert-butyl-4-methylphenol and an internal standard of 2H5-IAA (100 pg·mg−1 FW). The crude extract was concentrated under nitrogen gas. The aqueous residue was purified by solvent fractionation. The partially purified IAA samples were analyzed by liquid chromatography–tandem mass spectrometry (LC-MS/MS) with multiple reaction monitoring in the negative ion mode on a model 4000 QTRAP LC/MS/MS System (AB Sciex, Tokyo, Japan). The “precursor m/z > product m/z” was 174.1 > 130.1 for endogenous IAA and 179.1 > 135.1 for the internal standard. IAA was analyzed on a high-performance (HP) LC column (Capcell Pak C18 column, 150 × 2 mm; Shiseido, Tokyo, Japan) with a gradient solvent system (20%–100% methanol/water containing 0.1% acetic acid) at a flow rate of 0.1 mL·min−1.
For abscisic acid (ABA) quantification, samples were extracted in 5 mL of methanol–water–acetic acid (90:9:1, v/v/v) containing an internal standard of 13C2-ABA. Following extraction, 17.5 mL of water was added, and the samples were centrifuged. Oasis HLB cartridges (Waters, Mississauga, ON, Canada) conditioned with methanol and equilibrated with methanol–water–acetic acid (MWA; 9:90:1, v/v/v) were loaded with the samples and washed through with MWA (9:90:1). ABA was eluted with 1 mL MWA (90:9:1) and collected in clean tubes, and 5 μL of each sample was separated by HPLC on a C18 column (150 mm × 2 mm, 5 μm; Shiseido) at a flow rate of 0.2 mL·min−1 using a binary solvent system comprising methanol (A) and water with 0.1% formic acid (B). The compounds were analyzed by MS/MS with multiple reaction monitoring in the negative ion mode. The precursor m/z > product m/z was 263 > 153 for endogenous ABA and 265 > 153 for the 13C2-ABA internal standard.
Each sample was analyzed twice and the means are shown in the Figures.
To clarify the relationship between the number of flowers and the number of retained fruits in a cluster, we counted the numbers of total, abscised, and retained fruits per cluster a month after flowering (Fig. S2). Among the self-thinning accessions, ‘Chojuro’ retained 75% of fruits and ‘Niitaka’ retained 50% in 2009. The non-self-thinning ‘Kosui’ retained 93% (Table 1). ‘Akiakari’ and 162-29 retained 65% of fruits, with SD levels of 12% to 14%. ‘Chojuro’ retained 1 to 9 fruits per cluster and ‘Niitaka’ retained 1 to 7, both with SD ≥ 20% (Fig. 2; Table 1). The wide ranges of the SDs may be due to variations in both the numbers of total flowers in a cluster and numbers of abscised flowers/fruits: ‘Chojuro’ produced 5 to 10 flowers per cluster and abscised 0 to 5 fruits, and ‘Niitaka’ produced 4 to 8 flowers and abscised 1 to 7 fruits (Fig. S2a, c, d). In contrast, 162-29 retained 3 to 6 fruits and ‘Akiakari’ retained 3 to 5 fruits in most clusters (Fig. 2b, e, f). They also produced narrower ranges of flowers and abscised fruits: 162-29 produced 6 to 9 flowers on most clusters and abscised 1 to 5 fruits, and ‘Akiakari’ produced 5 to 7 flowers and abscised 0 to 4 fruits (Fig. S2b, e, f). Although ‘Chojuro’ and 162-29 had lower percentages of retained fruits in 2008 than in 2009, the pattern was similar (Fig. 2a–c, e), along with a smaller SD in 162-29 than in ‘Chojuro’ (Table 1). The non-self-thinning ‘Kosui’ produced 5 to 11 flowers per cluster and retained most of them (Figs. 2g and S2g; Table 1).
Distributions of numbers of retained fruits per cluster one month after hand-pollination. (a, c) ‘Chojuro’, (b, e) 162-29, (d) ‘Niitaka’, (f) ‘Akiakari’, (g) ‘Kosui’. (a, b) 2008, (c–g) 2009. For example, ‘Chojuro’ retained 2 fruits per cluster in 4 clusters in 2008 (a).
We analyzed the flowering order and the number of abscised (or retained) fruits in a cluster to investigate the relationship between them. Except in ‘Akiakari’, all flowers in a cluster bloomed within 4 to 6 days of the first flower to bloom. In ‘Chojuro’ in both years and in ‘Niitaka’ in 2009, there was no apparent relationship between flowering order and retained flowers (Figs. 2a, c and 3a, c, d). In 162-29, earlier flowers tended to be retained in the cluster; in particular, in 2008, 100% of the first flowers were retained, while nearly 100% of the last flowers were abscised (Fig. 3b, e). The thinning tendency of ‘Akiakari’ was similar to that of 162-29 (Fig. 3f). The non-self-thinning ‘Kosui’ retained most flowers regardless of flowering order (Fig. 3g). Because the number of fruits in a cluster is relevant to the number of retained fruits in a cluster in apple (Bangerth, 2000), we calculated the correlation between the number of flowers and that of retained fruits in a cluster. In ‘Chojuro’ and ‘Niitaka’, in which the determination of fruit retention is independent of the flowering order, the number of retained fruits was not correlated with the number of fruits in a cluster (Table 2). In 162-29 and ‘Akiakari’, in which the determination of fruit retention depends on the flowering order, there was a weak correlation (Table 2).
Distributions of retained fruit ratios by order of flowering day (1st to 5th) one month after pollination. Order means the order of the flowering day in the cluster. The fruits from the flowers that opened on the same day that the first flower in the cluster opened were designated “1st”. (a, c) ‘Chojuro’, (b, e) 162-29, (d) ‘Niitaka’, (f) ‘Akiakari’, (g) ‘Kosui’. (a, b) 2008; (c–g) 2009. For example, ‘Chojuro’ (a) retained 62% of fruits which opened on the 1st day in 2008. Data are means of 3 or 4 successive days (except z n = 2). Error bars indicate ± SEM. Data were analyzed by one-way ANOVA. Bars with the same letter indicate not significantly different among days (among 1st to 4th days in ‘Akiakari’) by the Tukey–Kramer multiple comparison test (P < 0.05).
Relationship between total fruits number and the percentage of abscised fruits.
Since the accumulation of auxin just after fruit set and auxin export are suggested to be critical for abscission in apple fruits (Bangerth, 2000), we measured the levels of IAA, the most widely distributed natural auxin, in flowers/fruits and spurs (basal part of a cluster). As self-thinning accessions abscise flowers/fruits 1 to 3 weeks after the beginning of flowering, flowers/fruits at 6 to 21 DAP were collected. The fate of each fruit was unpredictable, in particular for young fruits. In 162-29, when the number of flowers was restricted to two per cluster, all flowers (more than 100) developed into fruit without thinning. Therefore, to clarify the difference between fruits to be abscised and those to be retained, some clusters were left intact to permit self-thinning, whereas others were hand-thinned to two flowers per cluster on the day of first flowering to avoid self-thinning. In the thinned clusters, both fruits were always retained. In flowers/fruits of 162-29 in 2008, IAA increased transiently at 8 DAP, and its content in thinned clusters was ~1.5 × that in unthinned clusters (Fig. 4a). In spurs, it was relatively higher at 14 DAP in thinned clusters (Fig. 4c). Because ABA was suggested to be a trigger for thinning in citrus (Takahashi et al., 1975), we also measured the levels of ABA in flowers/fruits and spurs. In flowers/fruits, ABA decreased gradually, and more so in thinned clusters (Fig. 4b). In spurs, it increased, more so in thinned clusters, and the pattern was the mirror image of that in flowers/fruits (Fig. 4d). However, the differences in ABA content between thinned and unthinned clusters were small in both flowers/fruits and spurs. Therefore, we focused on IAA and measured its levels in fruits of all accessions in 2009. In all self-thinning accessions except ‘Akiakari’, IAA increased transiently at 6 DAP (Fig. 5a–d), then decreased and stayed low (Figs. 4a and 5a–d). In ‘Kosui’, it decreased at 6 DAP and then increased transiently with a peak at 11 DAP (Fig. 5e).
Comparison of endogenous IAA and ABA levels in flowers/fruits and spurs in unthinned and thinned clusters of 162-29 in 2008. Thinned clusters were reduced to 2 flowers at 6 days after pollination (DAP). IAA and ABA levels were measured at 6, 8, 14, and 21 DAP.
Changes in levels of IAA in flowers/fruits from unthinned clusters of self-thinning and non-self-thinning accessions at 4, 6, 8, 11, 14, and 20 DAP in 2009. (a) ‘Chojuro’, (b) 162-29, (c) ‘Niitaka’, (d) ‘Akiakari’, (e) ‘Kosui’. Closed triangles show the days when average cumulative temperature after pollination exceeded 90°C.
The percentage of retained fruits per cluster was lower in 2008 than in 2009 in both ‘Chojuro’ and 162-29 (Table 1). Furthermore, 162-29 showed a transient increase in IAA at 8 DAP in 2008, but at 6 DAP in 2009, along with all self-thinning accessions except ‘Akiakari’ (Figs. 4a and 5). To explain the difference in the abscission rate and the timing of auxin accumulation, we investigated the possible effects of temperature, which affects the abscission rate in apple (Bangerth, 2000). The temperature in spring in the orchard was lower in 2008 than in 2009 (http://komefuji.s101.xrea.com/sekisan.html, data point 47646; Tables 3 and S1, S2). The average cumulative temperature (integrated average daily temperature) was < 80°C at 6 DAP and > 100°C at 8 DAP in 2008, but ~90°C and >120°C in 2009 (Table 3).
Cumulative average temperature on stated days after pollination (DAP).
In this study, using Japanese pear accessions ‘Chojuro’, ‘Niitaka’, 162-29, and ‘Akiakari’, we investigated the pattern of fruit self-thinning and attempted to identify its trigger. As shown by ‘Kosui’, not all Japanese pears regulate their fruit number by self-thinning. The pedigree suggests that the self-thinning trait is inherited from ‘Chojuro’ (Fig. 1). In apple, the mean number of fruits per inflorescence and the percentage of inflorescences with one fruit are influenced by the genotype rather than by the total number of inflorescences per branch (Celton et al., 2014). This suggests that the self-thinning trait is genetically transmitted in apple. To reveal the mode of inheritance of the trait in pear, a collection of accessions with the fruit self-thinning trait and genealogical analysis using their descendants will be required.
In ‘Chojuro’ and ‘Niitaka’, the retention position in the cluster and the number of retained fruits were random. In 162-29 and ‘Akiakari’, they were relatively uniform. These results suggest that, in Japanese pears, the pattern of the number of fruits per cluster depends on accession, as it does in apple (Fig. 2; Celton et al., 2014). Since in Japanese pear, the 3rd to 5th flowers from the outside develop into fruits of good quality, the other flowers are removed by hand after full bloom. Therefore, the number of self-thinned fruits per cluster (4-6 in 162-29 and 3-5 in ‘Akiakari’) and their position in the cluster in 162-29 and ‘Akiakari’ are appropriate because they retain fruits developed from the 3rd to the 5th flowers (Fig. 2). This ideal self-thinning behavior of 162-29 and ‘Akiakari’ is different from that of ‘Chojuro’ and ‘Niitaka’, despite their shared pedigree. The hypothesis of primigenic dominance (PD) suggests that earlier developing organs inhibit later developing ones (Bangerth, 1989). The thinning pattern in 162-29 and ‘Akiakari’ was similar to that in apple, namely, the earlier the flower opens, the more likely it is to be retained in the cluster (Fig. 3). In 162-29 and ‘Akiakari’, the total fruit number in a cluster also affected the number of abscised fruits (Table 2). This pattern is consistent with the PD hypothesis. Even so, in ‘Chojuro’ and ‘Niitaka’, there was no relationship between fruit abscission and the order of the flowering (pollination) day, and the number of abscised fruits in a cluster was independent of the total number of flowers (Fig. 3; Table 2). Therefore, the self-thinning pattern observed in ‘Chojuro’ and ‘Niitaka’ is unlikely to be regulated in a PD manner. Since unpollinated fruits also abscised and the number of retained fruits was highly variable (from 0 to the total number of flowers minus 1), it is possible that not all the pollinated flowers successfully fertilized due to unknown physiological mechanisms in ‘Chojuro’ and ‘Niitaka’. In summary, there are three patterns of fruit self-thinning in Japanese pears, i.e. no self-thinning, PD-type self-thinning, and non-PD-type self-thinning. The difference between the PD-type and non-PD-type patterns suggests the involvement of multiple genetic loci in fruit self-thinning in Japanese pear. Breeding of ideal self-thinning accessions will require identification of more accessions with the self-thinning trait and genetic identification of the mode of inheritance.
To find the trigger for fruit self-thinning, we measured endogenous IAA and ABA levels (Figs. 4 and 5). Application of ABA induced thinning of young fruit in ‘Bartlett’, a European pear (Pyrus communis L.; Greene, 2012). Whether the effect of ABA on young fruit thinning is direct or indirect has long been discussed (Bangerth, 2000). A recent study in ‘Bartlett’ tested the hypothesis that if ABA directly induced fruit abscission, an additive effect of ABA and shade would be observed (Einhorn and Arrington, 2018). Since the effect of shade was greater than that of ABA and ABA reduced the carboxylation rate, the authors concluded that the effect of ABA was indirect and the major role of ABA in pear self-thinning may be to cause a carbohydrate deficit. In our experiments, the ABA content decreased in flowers/fruits and increased in spurs at 8 DAP, and more so in thinned clusters (Fig. 4). The difference in ABA levels between thinned and unthinned clusters suggests that ABA may be involved in fruit abscission. However, since the difference was modest, ABA seems unlikely to be a trigger for self-thinning in Japanese pear.
In apple, the higher the ranking in PD hierarchy in the cluster, the more IAA is exported from the fruit (Bangerth, 2000), implying that fruits with higher ranking produce more IAA. In 162-29, the IAA content increased transiently in flowers/fruits at 8 DAP in both thinned and unthinned clusters, but more so in the thinned clusters, in which all flowers/fruits were retained at one month after pollination (Fig. 4). This result suggests that the flowers/fruits to be retained had more IAA than those to be abscised, since the thinned clusters contained solely retaining flower/fruit groups, while unthinned clusters contained both retaining and abscising flower/fruit groups. This conclusion is supported by the fact that IAA increased in 162-29 and most of the other fruit self-thinning accessions, but not in ‘Kosui’, a non-self-thinning cultivar (Figs. 4 and 5). These results suggest that the increase in IAA in fruits to be retained is a candidate trigger for determining self-thinning in Japanese pear. This conclusion is consistent with the observations in apple. Among the self-thinning cultivars, no increase in endogenous IAA at 6 DAP was observed in ‘Akiakari’. The other accessions used in this study flowered for about a week, but ‘Akiakari’ flowered for a month in 2009. As auxin is also involved in the promotion of flowering (Tromp, 2000), we consider that the homeostasis of auxin in ‘Akiakari’ may have differed from that in other accessions, at least in 2009, and so the increase in IAA may have been masked by other events such as flowering of neighboring clusters.
The endogenous IAA level in 162-29 peaked at 8 DAP in 2008, but at 6 DAP in 2009. This difference suggests that self-thinning is also affected by factors other than the number of days after flowering. We concluded that the difference was an effect of temperature during self-thinning, which was higher in 2009 than in 2008 (Tables 3 and S1, S2). For example, the cumulative average temperature for ‘Chojuro’ from flowering day to 6 DAP was 79.1°C in 2008, but 95.5°C in 2009. In apple, a higher nighttime temperature stimulates fruit growth and increases the fruit abscission rate (Bangerth, 2000), while our data suggest that more fruits were abscised in 2008 than in 2009, even though the cumulative average temperature was higher in 2009 than in 2008. Therefore, using past data collected by the Japan Meteorological Agency (https://www.jma.go.jp/jma/index.html) the nighttime temperature over the test period was analyzed. As a result, the average nighttime temperatures during the test periods of “Chojuro” were 6.7 ± 1.8°C in 2008 and 8.3 ± 1.4°C in 2009 and those of 162-29 were 7.6 ± 1.6°C in 2008 and 7.5 ± 1.2°C in 2009; these were not significantly different between 2008 and 2009. The results suggest that the nighttime temperature was not the reason for the difference in the abscission rate in 2008 and 2009. In apple, aminovinylglycine (AVG) reduced the fruit abscission at high nighttime temperatures (Kondo and Takahashi, 1987). As AVG can inhibit auxin biosynthesis (Soeno et al., 2010), it may also inhibit fruit abscission induced by high nighttime temperatures by reducing auxin biosynthesis. Kondo and Takahashi (1987) also demonstrated that shading enhances abscission. Since shading induces auxin biosynthesis (Tao et al., 2008), auxin biosynthesis may be involved in abscission. In ‘Kosui’, a non-self-thinning cultivar, the level of endogenous auxin in fruits decreased at 6 DAP, at the time when it increased in self-thinning cultivars except ‘Akiakari’ (Fig. 5), implying that the transient increase in endogenous IAA is an event specific to fruit self-thinning cultivars.
Chemical thinning is used as a substitute for laborious hand thinning, but the response to chemical thinning is inconsistent and depends on the accession (Wertheim, 2000; Gonkiewicz et al., 2011). In this study, we found three patterns of fruit self-thinning, i.e. no self-thinning, PD-type-thinning, and non-PD-type-thinning. If the thinning pattern affects the response to a thinning agent, the response to chemical thinning can be estimated in part from the thinning pattern of the accession. Our data imply that the temperature-dependent increase in auxin is a trigger for fruit self-thinning in pear. However, the use of an auxin such as 1-naphthaleneacetic acid (NAA) as a thinning agent is not practical because of its multiple effects. Considering the stability of NAA and its molecular mechanism of transport in planta (Dunlap et al., 1986; Petrášek and Zažímalová, 2006), its effects on fruits will be different from that of endogenous IAA. Several auxin biosynthesis inhibitors have been reported (He et al., 2011; Kakei et al., 2015; Narukawa-Nara et al., 2016; Tsugafune et al., 2017). In this paper, we suggest that the relative abundance of endogenous auxin is important for determining retained fruits. Conversely, fruits with low auxin accumulation will be abscised, suggesting that auxin biosynthesis inhibitors could be considered as candidate thinning agents.
We thank Dr. Shozo Fujioka for technical assistance and critical comments on the manuscript. This research was supported by a grant from the Ministry of Agriculture, Forestry, and Fisheries of Japan.
