ORIGINAL ARTICLES
Differences in Water and Assimilate Fluxes in Tomato Fruits among Cultivars, and Relationships with Fruit Yield and Soluble Solids Content
2018 Volume 87 Issue 2 Pages 229-235
Details
2018 Volume 87 Issue 2 Pages 229-235
Fruit growth represents the balance between material influxes via xylem and phloem and efflux by transpiration via the stomata of the calyx and cuticle of fruit, which determines the yield and soluble solids content (SSC). Knowledge of these factors is important for the production of high-SSC tomato fruit, but no physiological indicator is available to allow prediction of fruit yield and SSC for breeding and crop production purposes. To identify indicators, we grew Japanese, Dutch, Japanese × Dutch, and high-SSC cultivars and sought correlations of the fluxes to fruit with yield and SSC. To estimate the contributions of the xylem, phloem, and transpiration fluxes to fruit weight increase, we measured 2-day growth rates of intact, detached, and heat-girdled (peduncle steamed for 90 to 120 s) fruits treated at 14, 28, or 42 days after flowering (DAF). Xylem influx was much lower in the high-SSC cultivar than in the others. Phloem influx was lower in the Dutch and hybrid cultivars at 28 DAF. Transpiration efflux was greater in the Japanese cultivar at 42 DAF. Fruit growth rate at 14 DAF was positively correlated with yield, and phloem influx per fruit weight increase at 14 and 28 DAF was positively correlated with SSC. These results show how the xylem, phloem, and transpiration fluxes of fruit can predict fruit yield and SSC. This information will help the production and breeding of high-SSC fruit.
Fruit yield is one of the most important factors in greenhouse production of tomatoes (Solanum lycopersicum L.). To increase it, many studies on breeding (Higashide and Heuvelink, 2009; van der Ploeg et al., 2007) and climate control (Bakker, 1990; Cockshull et al., 1992; De Koning, 1990; Drais et al., 2001; Heuvelink, 1999) have been carried out. Soluble solids content (SSC) is an important factor for fruit quality and commercial value. In Japan, SSC is often given more emphasis than fruit yield. Japanese tomato cultivars tend to have high SSC (Higashide et al., 2012), and SSC can be further increased through the use of osmotic stress caused by saline treatment, although the yield is decreased (Cuartero and Fernández-Muñoz, 1999; Ehret and Ho, 1986; Saito et al., 2006). In recent years, cultivars with inherently very high SSC have been released (Yamada et al., 2010). However, no physiological indicator of high SSC is available for use in production and breeding, and the related physiological processes are not fully understood (Saito et al., 2008).
Fruit SSC represents the balance between material influxes via the xylem and phloem and efflux via transpiration (Leonardi et al., 1999). Because assimilates are translocated only via the phloem, the proportion of phloem influx determines the fruit dry matter ratio, which strongly affects fruit SSC (Ho et al., 1987; Leonardi et al., 1999). In addition, fruit weight consists of the sum of these fluxes. Therefore, the values of these fluxes could be important indicators of high-SSC and high-yield fruit for both breeding and crop production purposes. If they could be controlled, fruit yield and SSC could also be controlled to a degree.
Some information on the characteristics of fluxes in tomato fruits is already available, having been evaluated through estimation of the calcium accumulation in fruit (Ho et al., 1987), comparison of growth between heat-girdled, detached, and intact fruits (Araki et al., 2004; Guichard et al., 2005; Hossain and Nonami, 2010; Kitano and Araki, 2001), and monitoring of water flow by nuclear magnetic resonance imaging (Windt et al., 2009). In this study, we used the heat-girdling method, heating or steaming the peduncle to kill phloem cells and restrict the phloem influx to fruit because this has been used to estimate transpiration and there are many reports in fruit trees such as Vitis vinifera L. (Greenspan et al., 1994, 1996; Lang and Thorpe, 1989), Malus pumila L. (Lang, 1990), Prunus persica (L.) Batsch (Fishman et al., 2001; Morandi et al., 2014), and P. avium L. (Brüggenwirth et al., 2016). In addition, systematic errors are very small in the short term with a statistically sufficient sample, and this is the only method that allows the estimation of xylem and phloem influx and transpiration in the greenhouse (Fishman et al., 2001) using the heat-girdling method because it is available for greenhouse tomato production and can evaluate many fruits in a short time.
The objective of this study was to explain the difference in fruit yield and SSC among some cultivars by the proportion of the fluxes via the xylem, phloem, and transpiration. This information can help high fruit yield and SSC tomato breeding and production.
We tested four tomato cultivars: ‘CF Momotaro York’ (Takii Seed, Kyoto, Japan), ‘Managua RZ’ (Rijk Zwaan, De Lier, The Netherlands), ‘Ringyoku’, and ‘DR03-103’ (Nippon Del Monte Agri, Tokyo, Japan). ‘CF Momotaro York’ is a common Japanese cultivar with a high SSC. ‘Managua RZ’ is a Dutch cultivar with a low SSC. ‘Ringyoku’ is an F1 hybrid derived from a cross between two pure lines that are progeny of the high-yield Dutch F1 cultivar ‘Geronimo’ (De Ruiter Seeds, Bergschenhoek, The Netherlands) and the Japanese F1 cultivar ‘Momotaro 8’ (Takii Seed) bred by the Institute of Vegetable and Floriculture Science, NARO (Matsunaga and Saito, 2017). ‘DR03-103’ has a higher SSC than popular Japanese cultivars.
Seeds were sown in 72-cell trays filled with commercial soil mix (Best Mix; Mitsubishi Plastics Agri Dream, Tokyo, Japan), and seedlings were grown for 3 weeks in a nursery chamber (Nae-Terrace; Mitsubishi Plastics Agri Dream) with day/night air temperatures of 25/18°C and 16 h of day length. On 19 September 2015, seedlings were transplanted into rockwool slabs in a plastic house (7.2 m × 14.2 m, Tsukuba, Japan) at a density of 4.17 plants·m−2. All plants received OAT-SA nutrient solution (OAT Agrio, Tokyo, Japan: 247 (μg·g−1) N, 105 P2O5, 480 K2O, 230 CaO, 60 MgO, 0.75 MnO, 1.1 B2O3, 2.3 Fe, 0.03 Cu, 0.09 Zn, 0.03 Mo) with electrical conductivity adjusted to 2.0 dS·m–1. Between 0.3–1.2 L·day−1 of the solution was supplied to each plant depending on growth stage. The cultivars were arranged in a randomized block design (3 treatment dates × 3 treatments in a block) with 8 replications. To achieve uniform fruit set on each truss and to promote fruit growth, we sprayed 80 μM 4-chlorophenoxyacetic acid after three flowers had opened on each truss. Plants were pinched after flowering of the 4th truss to leave 2 leaves above the 4th truss. The experiment ended on 20 January 2016. The plastic house was ventilated when the air temperature exceeded 26°C, and heated when it fell below 14°C. During the experiment, the air temperature averaged 18.1°C.
Evaluation of fluxes in fruitsWe used the 1st truss of each plant to investigate fluxes in fruits. Each truss, bearing 3 fruits, was treated at 14, 28, or 42 days after flowering (DAF, determined when 2 flowers had opened). To restrict both xylem and phloem influxes, the truss was detached from the stem, and the cut surface of the truss was coated with vaseline to avoid water loss. To restrict only phloem influx, we heat-girdled the peduncle of the truss by steaming it at approximately 95°C with a commercial steam cleaner (STM-415; Irisohyama, Sendai, Japan). According to our preliminary study the steaming time was set at 90 s (at 14 DAF) or 120 s (at 28 and 42 DAF). The peduncle was temporarily wrapped with aluminum foil for both physical protection and good heat transfer. Intact trusses were used as a control. Fluxes were calculated as differences in fruit growth rates between treatments as described below.
To estimate the fruit growth rate, we measured fruit diameters at the equator twice at 90° with digital calipers (ABS Digimatic Calipers CD-15APX; Mitutoyo Corporation, Kawasaki, Japan) just before treatment and used the average as the diameter. After 2 days, fruit diameter was measured as well. After that, we weighed the fruits and calculated regressions between diameter and weight. Fruit weights of each treatment and cultivar were estimated from the regression line of intact fruits on each treatment date because there was no difference among treatments (Fig. 1), and the normalized growth rate (G: g·g−1·day−1 per truss) was calculated as:
Correlations between cube of fruit diameter and fruit weight of different cultivars at successive developmental stages. A, B, C, and D are ‘CF Momotaro York’, ‘Managua RZ’, ‘Ringyoku’, and ‘DR03-103’, respectively. Values were measured 2 days after treatment. Regression lines were drawn with data of intact fruits. DAF: days after treatment. * Significant correlation at P < 0.05 (n = 24).
Xylem influx (X: g·day−1 per truss), phloem influx (P), and transpiration (T) of fruit were calculated according to Guichard et al. (2005) and Lang and Thorpe (1989) as:
To confirm that heat-girdling limited phloem influx, we observed peduncle tissues 2 days after treatment. Cross-sections immersed in 0.1% evans blue for 10 min to stain dead cells and then rinsed well were observed under a stereoscopic microscope (SZ61; Olympus, Tokyo, Japan) to compare damage to tissues between intact and heat-girdled peduncles.
Measurement of fruit yield and SSCRed-ripe fruits from the 2nd to 4th trusses were harvested and weighed. In addition, one control fruit of each cultivar was chosen on each harvest day for measurement of SSC (percentage of fresh weight) with a digital refractometer (PR-101; Atago, Tokyo, Japan).
Statistical analysesAll statistical analyses were carried out with JMP 11.0 software (SAS Institute Japan, Tokyo, Japan). We tested the significance of correlation coefficients of regressions for fruit diameter cubed against weight (n = 24). Differences in fruit fluxes, weight, and SSC among cultivars were detected by Tukey-Kramer’s multiple test (n = 7–8; 3 plants per replication). Differences in phloem influx per fruit weight increase among cultivars were tested by the Steel–Dwass nonparametric test (n = 7–8; 9 plants per replication) because data were not normally distributed. Effects of parameters on fruit yield and SSC were tested by regression analysis (n = 4; each value represents each cultivar’s average). Significant effects were reported at P < 0.05.
Staining with evans blue confirmed that the heat-girdling treatment limited phloem influx. In the intact peduncles of all cultivars, only xylem cells were stained, indicating cell death. In heat-girdled peduncles, phloem cells also were stained in all cultivars at all stages (Fig. 2).
Cross-sections of intact or heat-girdled peduncles stained with evans blue 2 days after treatment at 28–30 DAF. Arrows indicate phloem cells.
The truss weight of ‘DR03-103’ was significantly lower than the other 3 cultivars at all stages (Table 1). The increase in fruit weight of ‘DR03-103’ was significantly lower than the other 3 cultivars at 14–16 DAF. That of ‘Ringyoku’ was lower than that of ‘CF Momotaro York’ at 28–30 DAF, and that of ‘DR03-103’ was lower than that of ‘Managua RZ’ at 42–44 DAF.
Two-day fruit weight increases at successive fruit developmental stages.
The xylem influx of ‘DR03-103’ was significantly lower than those of the other 3 cultivars at all stages (Fig. 3A). There were no significant differences in phloem influx among cultivars at 14–16 or 42–44 DAF, but those of ‘CF Momotaro York’ and ‘DR03-103’ were significantly greater than those of ‘Managua RZ’ and ‘Ringyoku’ at 28–30 DAF (Fig. 3B). There were no significant differences in fruit transpiration at 14–16 DAF (Fig. 3C). That of ‘DR03-103’ was significantly lower than the other 3 cultivars at 28–30 DAF, and that of ‘CF Momotaro York’ was significantly greater than that of ‘DR03-103’ at 42–44 DAF.
(A) Xylem influx, (B) phloem influx, and (C) transpiration in 1st trusses of each cultivar at successive developmental stages. Error bars represent means ± SE (n = 7–8). Bars topped with the same letter at each stage are not significantly different at P < 0.05 by Tukey–Kramer’s test.
Phloem influx per fruit weight increase of ‘DR03-103’ was significantly greater than the other 3 cultivars at 14 DAF and than ‘Ringyoku’ and ‘Managua RZ’ at 28–30 and 42–44 DAF; that of ‘CF Momotaro York’ was greater than that of ‘Managua RZ’ at 42–44 DAF (Fig. 4).
Phloem influx per fruit weight increase of each cultivar at successive developmental stages. Error bars represent means ± SE (n = 7–8). Bars topped with the same letter at each stage are not significantly different at P < 0.05 by Steel–Dwass nonparametric test.
Total fruit yield per plant did not differ significantly among cultivars, but individual fruit weight decreased significantly in the order of ‘Ringyoku’ > ‘CF Momotaro York’ > ‘Managua RZ’ > ‘DR03-103’ (Table 2). The SSC of ‘DR03-103’ was the highest.
Fruit yield and soluble solids content.
Regression analyses showed that fruit yield was strongly positively correlated with fruit weight increase at 14–16 DAF, and SSC was correlated with phloem influx per fruit weight increase at 14–16 and 28–30 DAF (Fig. 5).
(A) Correlation between fruit yield and fruit weight increase. (B) Correlation between soluble solids content and phloem influx per fruit weight increase. Each value represents each cultivar’s average. * Significant correlation at P < 0.05 (n = 4).
Restricting phloem influx by heat-girdling aims to kill phloem cells. Microscopic observations in this study showed clearly that it killed the phloem cells and thus inhibited influx (Fig. 2). Most similar studies use linear variable displacement transducers or similar for the continuous measurement of fruit diameter to estimate fruit weight or volume (Araki et al., 2004; Brüggenwirth et al., 2016; Greenspan et al., 1994, 1996; Guichard et al., 2005; Kitano and Araki, 2001; Lang, 1990; Morandi et al., 2014). In this study, however, we used digital calipers (Hossain and Nonami, 2010) because we needed to measure many fruits (>100) in a short time. Our results show that fruit weight could be estimated with high accuracy from diameters measured in this way (Fig. 1). However, inaccuracies in some measurements were apparent (e.g., negative transpiration of ‘DR03-103’ at 28–30 DAF; Fig. 3C), so it is important to ensure that measurement errors are minimized.
Significant differences among cultivars in xylem and phloem influx and transpiration were evident (Fig. 3, 4). In particular, the lower xylem influx in ‘DR03-103’ (Fig. 3A) could explain the high SSC (Table 2). Phloem influx tended to be high in the Japanese cultivars ‘CF Momotaro York’ and ‘DR03-103’. Although we tested only 2 cultivars, this may be a characteristic of Japanese cultivars, most of which have been bred with a focus on flavor instead of high yield over the past several decades (Higashide et al., 2012). Fruit transpiration, which also affects SSC, was high in ‘CF Momotaro York’. The high xylem influx and low phloem influx of the Dutch cultivar ‘Managua RZ’ can explain its low SSC. The intermediate SSC value of the hybrid cultivar ‘Ringyoku’ between those of ‘CF Momotaro York’ and ‘Managua RZ’ is consistent with the hybrid’s derivation from Japanese and Dutch parents (Matsunaga and Saito, 2017).
Xylem and phloem influxes and their ratio differed between our results and previous reports. For example, water influx to fruit derived from phloem exceeded 80% (Ho et al., 1987), was approximately 75% (Guichard et al., 2005), 60% (Araki et al., 2004), and 25% (Windt et al., 2009), respectively. These differences could be due to differences in evaluation methods, cultivars, and environmental conditions, so although the comparisons within this study are effective, comparison among reports may mean little.
Bertin (2005) reported that fruit growth rate is maximum at 10 to 15 DAF, although it is affected by temperature. The positive correlation of fruit yield with fruit weight increase at 14–16 DAF (Fig. 5A) suggests that it would be suitable to represent total fruit growth. Fruit SSC is affected by phloem influx and the sugar content in the phloem sap because fruit assimilates are derived only from phloem sap. SSC was positively correlated with phloem influx per fruit weight increase at 14–16 and 28–30 DAF (Fig. 5B). Although fruit respiration also affects fruit SSC, it can be neglected because it is much smaller than the dry matter increase (Grange and Andrews, 1995). Although there is no information about differences in the sugar contents of phloem sap among cultivars, the contents of soluble sugars, starch, and organic acids in phloem sap change with source to sink ratios (Jan and Kawabata, 2011), so they could differ among cultivars. Therefore, SSC was not completely explained by the phloem influx per fruit weight increase at 14–16 and 28–30 DAF, but the above can be important parameters to predict fruit SSC.
In conclusion, we revealed fluxes of xylem, phloem, and transpiration during tomato fruit growth, which differed among cultivars, and identified parameters correlated with fruit yield and SSC. This knowledge may be useful to achieve high yield and SSC in tomato breeding. Since the parameters can be measured before fruit ripening, yield and SSC can be estimated before harvest and the time for selection in breeding can be reduced.
We thank Mr. Tetsuya Saito for helping with plant management and investigations in the greenhouse.
