Involvement of berry hormonal content in the response to pre- and post-veraison water deficit in different grapevine (Vitis vinifera L.) cultivars
Abstract
Background and Aims
The application of deficit irrigation to grapevines modifies the hormonal status of berries, but little information about the influence of berry hormones on phenological sensitivity to water deficit is available. Therefore, the aim of this research was to assess the involvement of berry hormonal status in fruit composition in response to regulated deficit irrigation applied during different phenological stages in two grapevine cultivars.
Methods and Results
The study was carried out on fruiting cuttings of two cultivars of Vitis vinifera L., Tempranillo and Graciano. Treatments were: (i) early water deficit from fruitset to onset of veraison (early deficit); (ii) late water deficit from onset of veraison to harvest (late deficit); and (iii) plants regularly irrigated (Control). Both early water deficit and late water-deficit strategies modified evolution of indole-3-acetic acid, abscisic acid, salicylic acid and jasmonic acid, which was related to changes in berry size, increases in phenolic substances and accumulation of amines.
Conclusions
Differential sensitivity of Tempranillo and Graciano grapevines to seasonal water stress was mediated, at least in part, by alterations in hormonal status of berries at the time of water stress imposition.
Significance of Study
This study relates interspecific differences in the sensitivity of seasonal water-deficit irrigation to changes in the endogenous hormonal status of berries.
Introduction
Population growth, economic development, environmental demands and climate change converge into a scenario of water scarcity worldwide and therefore, water supply may constrain grapevine production and wine composition (Ortega-Farias et al. 2012). In this context, regulated deficit irrigation (RDI) has emerged as a potential strategy to control vegetative development and reduce berry size, improve cluster microclimate, increase water use efficiency and the concentration of sugar and phenolic substances in several winegrape cultivars and under different edaphoclimatic conditions (McCarthy et al. 2002, Kriedemann and Goodwin 2003, Chaves et al. 2010, Romero et al. 2010). In an RDI system, water restriction is applied during determined phenological stages when the effect on fruit growth and composition are neutral or positive, while keeping vineyard vigour in balance with potential production (Greven et al. 2005, Pellegrino et al. 2006, Ruiz-Sánchez et al. 2010, Sofo et al. 2012). It is critical, however, to understand responses to the timing of irrigation so that appropriate deficit irrigation strategies are applied that achieve adequate yield and berry composition. It has been reported that the effect of RDI depends on the vine phenological stage and the severity of the stress. Indeed, a pre-veraison water deficit is known to reduce shoot growth and berry early development, which results indirectly in earlier grape ripening (Castellarin et al. 2007a,b) and increased accumulation of phenolic substances (Deluc et al. 2009, Ollé et al. 2011, Santesteban et al. 2011). In this strategy, irrigation is resumed after veraison in order to maintain leaf activity for promotion of sugar synthesis and translocation into the berries (Wample and Smithyman 2002). In contrast, a post-veraison water deficit reduced berry size but increased the proportion of the fresh mass of the whole berry represented by seeds and skin, which results in a higher content of skin phenolics (Intrigliolo and Castel 2010, Ollé et al. 2011). Previous reports have shown that berry composition is characterised by a substantial differential sensitivity to water stress that is also dependent on grapevine cultivar. For instance, in Tempranillo, a pre-veraison water deficit reduced berry size and changed berry composition (Girona et al. 2009), but a post-veraison water shortage impaired berry sugar accumulation because of the detrimental effect of water stress on photosynthesis (Intrigliolo et al. 2012). Moreover, in Chardonnay, a post-veraison water limitation increased the concentration of total amino acids but decreased wine chemical and sensory attributes (Basile et al. 2012). In contrast, a post-veraison water deficit applied to Cabernet Sauvignon increased the concentration of anthocyanins and phenolic substances (Basile et al. 2011).
A large body of evidence has shown that grape berry growth and ripening are controlled by plant hormones (Davies and Böttcher 2009). In addition, under water-deficit conditions, our previous findings have shown that alterations to berry hormone dynamics conditioned berry composition and the profiles of phenolic substances and nitrogen compounds (Niculcea et al. 2013). The implications, however, of endogenous changes to berry hormones in response to water deficit imposed at different phenological stages are poorly understood. Therefore, the aim of this work was to assess changes in berry hormonal status in response to two RDI strategies applied from fruitset to the onset of veraison or from the onset of veraison to harvest, and their possible implications for berry composition in two original Spanish winegrape cultivars, Tempranillo and Graciano. Because viticulture practices, such as nitrogen supply (Goutu et al. 2012) and irrigation schedules (Bover-Cid et al. 2006, Basile et al. 2012, Niculcea et al. 2013), can modify the concentration of nitrogen compounds in the berry, special attention was paid to the effect of both RDI regimes on berry nitrogen profiles. Potted vines facilitated the imposition of a similar water stress procedure in both cultivars.
Material and methods
Plant material
Dormant Vitis vinifera (L.) cuttings of cvs Tempranillo and Graciano 400–500 mm long and 15–20 mm in diameter were selected in the winter of 2011. The cuttings were propagated by a technique that ensured that the formation of adventitious roots preceded budburst using steps originally outlined in Mullins (1966) with some modifications described in Ollat et al. (1998) and Antolín et al. (2010). Cuttings were rooted in a heat bed (25°C) inside a cool room (5°C) for 30 days. Rooted cuttings were planted in 10-L plastic pots containing a soil peat (1:1, v/v) potting mix. After rooting, cuttings were transferred to a greenhouse with a 25/20°C and 70/80% relative humidity (RH) (day/night) regime. They were illuminated for 15 h with natural daylight supplemented with metal halide lamps (POWERSTAR, HQI-TS, 400W/D, PRO, Osram, Augsburg, Germany), providing a minimum photosynthetic photon flux density (PPFD) of 350 μmol/m2·s at the level of the inflorescence. Budbreak occurred after 1 week under these conditions. Careful control of vegetative growth before flowering improves the partitioning of stored carbon toward the roots and the reproductive structures. Thus, only a single flowering stem was allowed to develop on each plant during growth. After berry set (end of May), growth conditions in the greenhouse were changed to a 25/15°C and 60/80% RH (day/night) regime with a PPFD of 600 μmol/m2·s at the level of the inflorescence. A nutrient solution provided mineral nutrition in accordance with viticulture requirements (Ollat et al. 1998).
Samples were obtained from berries collected at four stages of development: 25–30 days after anthesis (DAA), corresponding to pea-size berries (7 mm diameter) (Eichhorn and Lorenz (E–L) growth stage 31, pea size) (Coombe 1995); 60–70 DAA, when berries began to colour and enlarge (approximately 9°Brix) (E–L 35 stage, veraison); 70–85 DAA, red berries not quite ripe (approximately 16°Brix) (E–L 37 stage, almost ripe); and 100–110 DAA, corresponding to commercially ripe berries (approximately 22°Brix) (E–L 38 stage, harvest). There were five replicates for each treatment and sampling–time combination.
Water treatments
The irrigation programs for the different water supply treatments are summarised in Table 1. Two RDI strategies were compared with a complete irrigation as a Control. Treatments were applied after fruitset (E–L stage 27) (Coombe 1995). In the Control treatment, pots were maintained at 80% of pot capacity. In the RDI treatments, plants received 50% of the water given to Control plants from fruitset to onset of veraison (early deficit, ED) or from onset of veraison to harvest (late deficit, LD). The resulting soil conditions during the experiments are shown in Figure 1. Volumetric soil water content was monitored with an EC 5 water sensor (Decagon Devices, Inc., Pullman, WA, USA) placed within each pot. Pot capacity was previously assessed by determining water retained after free-draining water had been allowed to pass through the holes in the bottom of the pot. The surface of the plant containers was covered with quartz stones during the experiments to avoid water loss because of evaporation. Watering was performed with nutrient solution or deionised water in order to supply the different treatments with the same amount of nutrients during water deficit.
Effect of three irrigation treatments applied during berry development and ripening on the soil water content recorded from fruitset to harvest for (a,b) Tempranillo and (c,d) Graciano in pots. The irrigation treatments were full irrigation (Control) (●); (a,c) early water deficit from fruitset to onset of veraison (early deficit) (□); and (b,d) late water deficit from onset of veraison to harvest (late deficit) (∆). Values represent means (n = 5). Arrow indicates the onset of veraison.
| Treatment | Days after anthesis | ||||
|---|---|---|---|---|---|
| Tempranillo | 10 | 30 | 60 | 70 | 98–105 |
| Graciano | 10 | 25 | 70 | 85 | 106–110 |
| Fruit set | Pea size | Veraison | Almost ripe | Harvest | |
| Control (%) | 80 PC | 80 PC | 80 PC | 80 PC | 80 PC |
| ED (%) | 40 PC | 40 PC | 80 PC | 80 PC | 80 PC |
| LD (%) | 80 PC | 80 PC | 40 PC | 40 PC | 40 PC |
- ED, early season deficit irrigation from fruitset to veraison; LD, late season deficit irrigation from veraison to harvest; PC, pot capacity.
Predawn leaf water potential (Ψ) was measured with a SKYE SKPM 1400 pressure chamber (Skye Instruments Ltd, Llandrindod, Wales) on five fully expanded leaves per treatment at each sampling date just prior to irrigation. Ninety to 100 berries from each treatment (45–50 berries per biological replicate) were counted, weighed and frozen at −80°C for further analysis.
Phenolic substances composition
Total and extractable anthocyanins and phenolic substances were calculated according to the procedure described by Saint-Cricq et al. (1998), a method widely used in wineries that has been proven as a good tool for estimating wine colour (Kontoudakis et al. 2010). Briefly, two samples of the unfiltered homogenate were macerated for 4 h at pH 1 (HCl) and pH 3.2 (tartaric acid), respectively. Once maceration was completed, the macerated samples were centrifuged, and the supernatants were used for the following determinations: the concentration of phenolic substances was measured spectrophotometrically in the supernatant obtained after maceration at pH 3.2 at 280 nm; and anthocyanins were determined in both supernatants according to Ribéreau-Gayon and Stonestreet (1965) using absorbance at 520 nm. The cellular extractability was calculated from both data as described in Salazar-Parra et al. (2010).
Analysis of nitrogen-containing compounds
Aminoenone derivatives were obtained by reaction of o-phthalaldehyde. Samples were previously diluted in 0.1 mol/L HCl and then were passed through a 0.22-μm particle filter. The nitrogen-containing compounds were separated with an ACE (Advanced Chromatography Technologies Ltd, Aberdeen, Scotland) high-performance liquid chromatography (HPLC) column 4 μm particle size (150 mm × 3.9 mm), with temperature controlled at 16°C. Under these conditions, 21 compounds were separated, identified and quantified in a single injection: 19 amino acids and 2 biogenic amines. Target compounds were identified according to the retention times and ultraviolet–visible spectroscopy (UV–Vis) spectral characteristics of the derivatives of the corresponding standards, and they were quantified using the internal standard method. Analyses were performed in triplicate.
Hormone determination in berries
The concentration of abscisic acid (ABA), indole-3-acetic acid (IAA), salicylic acid (SA) and jasmonic acid (JA) in berries was determined by HPLC coupled to tandem mass spectrometry following protocols from Durgbanshi et al. (2005) and Montoliú et al. (2009) as described elsewhere (Niculcea et al. 2013). Frozen berries were ground and then 0.5 g of powdered tissue was extracted in ultrapure water for 3 min. Before extraction, 5 μL of a mixture of internal standards containing 100 ng of [2H6]-ABA, 5 ng of [2H2]-IAA, 100 ng of [2H4]-SA and 100 ng of [2H6]-JA were added to assess recovery and matrix effects. After extraction and centrifugation (5000 g, 10 min) at 4°C, the pH of the supernatant was adjusted to 3.0 and partitioned twice against diethyl ether. The organic layers were combined and evaporated in a centrifuge vacuum evaporator (Jouan, Saint-Herblain, France). The dry residue was thereafter resuspended in a water : methanol (9:1) solution, filtered and injected into a Alliance 2695 HPLC system (Waters Corporation, Milford, MA, USA). Hormones were separated in a reversed-phase Kromasil 100 C18 column (100 × 2.1 mm, 5 μm particle size)) (Phenomenex, Torrance, CA, USA) using methanol and ultrapure water both supplemented with glacial acetic acid to a concentration of 0.05%. The mass spectrometer, a triple quadrupole system (Quattro LC, Micromass Ltd, Manchester, England), was operated in negative ionisation electrospray mode, and plant hormones were detected according to their specific transitions using a multiresidue mass spectrometric method.
Yield and fruit composition
Bunches from each treatment type were weighed and all berries from each bunch were separated and weighed to obtain yield. Berries were weighed individually; mean berry mass was determined; and berries were separated into skin and flesh. Berry volume was calculated using the formula for a sphere (4/3πr3, where r is the radius of the sphere). Leaf area was measured with a portable area meter (model LI-3000, Li-Cor, Lincoln, NE, USA). Ten berries from each treatment were collected, weighed and subsequently oven dried at 80°C until constant mass.
A sample of 25 berries was crushed for determination of total soluble sugars, pH and titratable acidity. Soluble solids were measured using a temperature-compensating refractometer (Zuzi model 315, Auxilab S. L., Beriàin, Spain) and were expressed as Brix. Must pH was measured with a pH meter (Crison Instruments, Barcelona, Spain) standardised to pH 7.0 and 4.0. Finally, titratable acidity was measured by titration with NaOH, and was expressed as g tartaric acid/L according to the methods of the Organisation Internationale de la Vigne et du Vin (Organisation Internationale de la Vigne et du Vin 2009).
Statistical analysis
Data recorded at harvest were subjected to a two-factor analysis of variance (ANOVA), and variance was related to the main treatments (cultivar and irrigation treatment) and to the interaction between them. Data for hormones were analysed by a three-factor ANOVA to partition the variance into the main effects and the interaction between cultivars, developmental stage and irrigation regime. Means ± standard errors (SE) were calculated and, when the F ratio was significant, least significant differences were evaluated by a Tukey's t-test by using the Statistical Package for the Social Sciences (SPSS) (SPSS Inc., Chicago, IL, USA) version 15.0 for Windows XP. All values shown in the figures are means ± SE.
Results
Plant water status and plant growth
The irrigation treatments caused significant differences in grapevine water status, as indicated by the decrease in predawn leaf Ψ measured in plants subjected to water deficit at pre-veraison (ED) or at post-veraison (LD) compared with well-irrigated plants (Figure 2a,c. During ED treatment, the predawn leaf Ψ values reached in Tempranillo and Graciano were around −1.1 and −1.2 MPa. In the LD treatment, the lowest value of predawn leaf Ψ was recorded after veraison with a minimal value of −1.35 and −1.4 MPa for Tempranillo and Graciano, respectively.
Effect of three irrigation treatments applied during berry development and ripening: full irrigation (Control) (
); early water deficit from fruitset to onset of veraison (early deficit) (
); and late water deficit from onset of veraison to harvest (LD) (
) on (a,c) the predawn leaf water potential (Ψ) and (b,d) plant leaf area of fruiting cuttings of (a,b) Tempranillo and (c,d) Graciano. At pea size and veraison, Control and late deficit had the same values because they were essentially the same treatment. Values represent means (n = 5). Different letters indicate a significant difference (P < 0.05) between treatments according to Tukey's test.
Tempranillo is characterised by a shorter reproductive cycle than Graciano (Table 1). In Tempranillo, the ED and LD treatments prolonged the time from veraison to berry maturity by 8 and 6 days, respectively. The effect of water-deficit irrigation on phenology, however, was less evident in Graciano. Imposition of any water-deficit period during the reproductive cycle reduced yield (Table 2) and vegetative growth, as shown by the marked decrease in leaf area obtained in both cultivars after submission to ED and LD treatments (Figure 2b,d). The balance between vine supply capacity and crop demand (i.e. crop load) expressed in terms of leaf area per yield was decreased in LD-treated plants of Tempranillo and in ED- and LD-treated plants of Graciano (Table 2).
| Cultivar | Irrigation | Yield (g/plant) | Crop load (cm2/g) | Berry mass (g/berry) | Berry volume (mm3) | Relative skin mass (% berry FM) |
|---|---|---|---|---|---|---|
| Tempranillo | Control | 186.61 a | 32.75 b | 0.93 ab | 928.36 a | 36.29 c |
| ED | 127.54 b | 20.82 bc | 0.85 ab | 578.96 b | 44.21 a | |
| LD | 136.90 b | 13.15 c | 0.86 ab | 916.13 a | 40.53 b | |
| Graciano | Control | 103.43 b | 54.26 a | 0.98 a | 711.54 b | 36.39 c |
| ED | 82.93 c | 22.21 bc | 0.75 b | 600.26 b | 37.45 bc | |
| LD | 85.05 c | 13.76 c | 0.88 ab | 635.81 b | 39.91 bc | |
| Cultivar (C) | *** | ** | ns | *** | ** | |
| Irrigation (I) | *** | *** | ** | *** | *** | |
| C × I | ns | ** | ns | ** | ** | |
- Values represent means (n = 5). Within each column, means followed by different letters are significantly different (P < 0.05) according to Tukey's test. FM, fresh matter. ns, ** and ***, not significant and significant at 1% and 0.1% probability level, respectively. ED, early season deficit irrigation from fruitset to veraison; LD, late season deficit irrigation from veraison to harvest.
The effect of water-deficit treatments on berry characteristics differed according to the cultivar analysed, and this differential pattern was emphasised by two-way ANOVA showing significant interaction between cultivar and irrigation treatment for most berry characteristics recorded (Table 2). Thus, in Tempranillo, ED treatment reduced berry volume and increased relative skin mass, but in Graciano, early water deficit decreased only berry mass. In contrast, LD treatment produced berries with high relative skin mass in Tempranillo, but no significant change in other berry traits were detected in Graciano.
Berry and must composition
Analysis of grape must at harvest indicate a significant difference between water regimes and cultivars assayed in this study (Table 3). In Tempranillo, berries from ED treatment have high total soluble solids (°Brix), high pH, low titratable acidity and low malic acid in comparison with those of berries from Control plants. In Graciano, ED-treated berries have high total soluble solids, low titratable acidity and low malic acid. By contrast, the LD treatment resulted in increased total soluble sugars in both cultivars. In Tempranillo, however, the ED and LD treatments improved accumulation of total phenolic substances compared with that of berries from well-watered plants, but no significant changes were detected for Graciano (Table 3). As a consequence, two-factorial ANOVA showed significant interaction between both factors for the concentration of phenolic substances evidencing that each cultivar responded in a different way to water-deficit treatments (Table 3).
| Cultivar | Irrigation | Total soluble solids (°Brix) | Must pH | Titratable acidity (g/L) | Malic acid (g/L) | Phenolic substances (mg/g) |
|---|---|---|---|---|---|---|
| Tempranillo | Control | 20.76 b | 3.86 b | 5.42 ab | 5.92 a | 1.79 b |
| ED | 22.37 a | 4.15 a | 4.23 c | 4.23 b | 3.04 a | |
| LD | 22.04 a | 4.03 ab | 4.85 bc | 5.10 a | 3.20 a | |
| Graciano | Control | 20.31 b | 3.37 c | 6.03 a | 5.47 a | 3.45 a |
| ED | 22.30 a | 3.59 c | 4.79 bc | 2.43 b | 3.40 a | |
| LD | 22.75 a | 3.51 c | 5.37 ab | 5.67 a | 3.73 a | |
| Cultivar (C) | ns | *** | *** | ns | *** | |
| Irrigation (I) | *** | *** | *** | *** | *** | |
| C × I | ns | ns | ns | ns | ** | |
- Values represent means (n = 5). Within each column, means followed by different letters are significantly different (P < 0.05) according to Tukey's test. ns, ** and ***, not significant and significant at 1% and 0.1% probability level, respectively. ED, early season deficit irrigation from fruitset to veraison; LD, late season deficit irrigation from veraison to harvest.
Under well-watered conditions, total and extractable anthocyanins were higher in Tempranillo than in Graciano berries, and the application of water-deficit irrigation resulted in different effects on each cultivar (Figure 3a,b). The ED procedure reduced the concentration of total anthocyanins in Tempranillo, but it was increased in Graciano. This treatment, however, significantly decreased the concentration of extractable anthocyanins in the two cultivars. The LD treatment did not produce any significant change in total or extractable anthocyanins in Tempranillo, but in Graciano there was a significant accumulation of anthocyanins. Both water deficit treatments also increased the cellular extractability with respect to well-irrigated plants, especially in Graciano (Figure 3c).
Effect of three irrigation treatments applied during berry development and ripening: full irrigation (Control) (
); early water deficit from fruit set to onset of veraison (early deficit) (
); and late water deficit from onset of veraison to harvest (LD) (
) on (a,d) total anthocyanins, (b,e) extractable anthocyanins and (c,f) cellular extractability measured at harvest in berries from fruiting cuttings of (a–c) Tempranillo and (d–f) Graciano. Values represent means (n = 5). Different letters indicate a significant difference (P < 0.05) between treatments according to Tukey's test.
Berry nitrogen compounds
Berries from Tempranillo have the highest concentration of total amino acids irrespective of the water treatment applied (Figure 4a). The main amino acids identified in berries are shown in Table 4. Berries of Tempranillo have a higher concentration of tyrosine, aspartic acid, asparagine, glutamic acid, glutamine, histidine, arginine and valine than those of Graciano berries. In addition, there were specific effects of water deficit on nitrogen compounds. Thus, in Tempranillo, the ED irrigation resulted in a high amino acid concentration because of accumulation of serine, histidine and arginine in comparison with that of berries from well-watered plants (Table 4). In Graciano, ED-treated berries, however, have a higher concentration of tryptophan, threonine, isoleucine, serine and alanine than that in well-watered berries. In contrast, the LD treatment did not alter total amino acid concentration with respect to that of well-watered plants in Tempranillo, but increased the amino acid concentration in Graciano (Figure 4a,c). In this cultivar, total amine concentration was strongly increased by ED and LD treatments but no significant change was detected in Tempranillo (Figure 4d,b). The main amine identified in berries was putrescine and to a lesser extent, methylamine (Table 4). Our data showed that the concentration of these compounds differed according to irrigation schedule and cultivar. Thus, two-factorial ANOVA showed a significant interaction between both factors.
Effect of three irrigation treatments applied during berry development and ripening: full irrigation (Control) (
); early water deficit from fruit set to onset of veraison (ED) (
); and late water deficit from onset of veraison to harvest (LD) (
) on (a,c) total amino acid and (b,d) amine concentration in berries from fruiting cuttings of (a,b) Tempranillo and (c,d) Graciano. Values represent means (n = 5). Different letters indicate a significant difference (P < 0.05) between treatments according to Tukey's test.
| Concentration (mg/g) | Cultivar (C) | Irrigation (I) | C × I | |||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Tempranillo | Graciano | |||||||||
| Precursor | Amino acid/amine | Control | ED | LD | Control | ED | LD | |||
| Amino acids | ||||||||||
| 3-Phosphoglycerate | Glycine | 0.0022 a | 0.0025 a | 0.0024 a | 0.0027 a | 0.0037 a | 0.0032 a | ns† | ns | ns |
| Serine | 0.0448 b | 0.0647 a | 0.0388 b | 0.0478 b | 0.0713 a | 0.0536 ab | ns | ** | ns | |
| Phosphoenolpyruvate | Tyrosine | 0.1577 a | 0.1666 a | 0.1321 a | 0.0422 b | 0.0427 b | 0.0440 b | *** | ns | ns |
| Tryptophan | 0.0189 a | 0.0276 a | 0.0211 a | 0.0127 b | 0.0180 a | 0.0207 a | * | ns | ns | |
| Phenylalanine | 0.0127 a | 0.0116 a | 0.0087 ab | 0.0069 b | 0.0090 ab | 0.0094 ab | ns | ns | ns | |
| Oxaloacetate | Aspartic acid | 0.0565 a | 0.0760 a | 0.0498 a | 0.0184 b | 0.0133 b | 0.0176 b | *** | ns | ns |
| Asparagine | 0.0291 a | 0.0454 a | 0.0379 a | 0.0051 b | 0.0055 b | 0.0082 b | *** | ns | ns | |
| Threonine | 0.0739 ab | 0.1081 a | 0.0806 a | 0.0603 b | 0.0861 a | 0.0938 a | * | * | ns | |
| Methionine | 0.0094 a | 0.0088 a | 0.0069 a | 0.0058 a | 0.0064 a | 0.0091 a | ns | ns | ns | |
| Isoleucine | 0.0163 a | 0.0210 a | 0.0143 ab | 0.0111 b | 0.0161 a | 0.0176 a | * | * | ns | |
| Lysine | 0.0059 ab | 0.0081 a | 0.0055 ab | 0.0038 b | 0.0050 ab | 0.0054 ab | * | ns | ns | |
| α-Ketoglutarate | Glutamic acid | 0.1154 a | 0.1557 a | 0.1280 a | 0.0479 b | 0.0544 b | 0.0545 b | *** | ns | ns |
| Glutamine | 5.2249 a | 7.1798 a | 5.5875 a | 2.2036 b | 3.2363 b | 4.2304 ab | *** | ns | ns | |
| Histidine | 0.0517 bc | 0.0949 a | 0.0652 ab | 0.0206 c | 0.0233 c | 0.0315 bc | *** | ns | ns | |
| Arginine | 0.7796 b | 1.0893 a | 0.7226 b | 0.3751 c | 0.5143 bc | 0.5859 bc | *** | * | ns | |
| γ-aminobutyric acid | 0.0611 b | 0.0575 b | 0.0526 b | 0.0804 a | 0.1033 a | 0.1155 a | *** | ns | ns | |
| Pyruvate | Alanine | 0.1327 a | 0.1589 a | 0.1139 ab | 0.0856 b | 0.1483 a | 0.1220 ab | ns | * | ns |
| Valine | 0.0212 a | 0.0199 a | 0.0249 a | 0.0034 b | 0.0026 b | 0.0039 b | *** | ns | ns | |
| Leucine | 0.0163 ab | 0.0210 a | 0.0143 ab | 0.0111 b | 0.0161 ab | 0.0176 ab | ns | ns | ns | |
| Amines | ||||||||||
| α-Ketoglutarate | Putrescine | 0.0003 c | 0.0002 c | 0.0003 c | 0.0002 c | 0.0004 b | 0.0005 a | *** | *** | *** |
| Methylamine | 0.00005 a | 0.00012 a | 0.00006 a | ND | ND | 0.00009 a | ** | * | *** | |
- Values represent means (n = 5). Within each column, means followed by different letters are significantly different (P < 0.05) according to Tukey's test. ns, *, ** and ***, not significant and significant at 5%, 1% and 0.1% probability level, respectively. ED, early season deficit irrigation from fruitset to veraison; LD, late season deficit irrigation from veraison to harvest; ND, not detectable.
Berry hormonal concentation
The concentration of ABA, IAA SA, and JA in berries is distinctly affected by the three factors considered, i.e. cultivar, development stage and water regime (Figure 5), and this was highlighted by a three-factor ANOVA analysis (Table 5). The factor ‘Cultivar’ significantly influenced the concentration of JA and SA (P ≤ 0.001) but the interaction between ‘Developmental stage’ and ‘Irrigation’ was highly significant for all berry hormones analysed (P ≤ 0.001). Results also showed that interaction between ‘Cultivar’, ‘Developmental stage’ and ‘Irrigation’ was significant for IAA and JA (P ≤ 0.001).
Effect of three irrigation treatments applied during berry development and ripening: full irrigation (Control) (
); early water deficit from fruit set to onset of veraison (ED) (
); and late water deficit from onset of veraison to harvest (LD) (
) on the the concentration of (a,e) abscisic acid (ABA), (b,f) indole-3-acetic acid (IAA), (c,g) jasmonic acid (JA) and (d,h) salicylic acid (SA) in berries from fruiting cuttings of (a–d) Tempranillo and (e–h) Graciano. At pea size and veraison, Control and LD had the same values because they were essentially the same treatment. Values represent means (n = 15). Different letters indicate a significant difference (P < 0.05) between treatments according to Tukey's test.
| Factors | ABA | IAA | JA | SA |
|---|---|---|---|---|
| Cultivar (C) | ns | ns | *** | *** |
| Developmental stage (D) | *** | *** | *** | *** |
| Irrigation (I) | *** | *** | *** | *** |
| C × D | ns | *** | *** | *** |
| C × I | *** | *** | *** | ns |
| D × I | *** | *** | *** | *** |
| C × I × D | ns | *** | *** | ns |
- ns and ***, not significant and significant at 0.1% probability level, respectively. ABA, abscisic acid; IAA, indole-3-acetic acid; JA, jasmonic acid; SA, salicylic acid.
In well-watered plants, the highest concentration of ABA was detected at veraison in both cultivars (Figure 5a,e). In the ED treatment, a significant increase in ABA took place earlier than in Control plants (at the pea size), reaching a maximum concentration at veraison. In the LD treatment, ABA accumulation lasted until at least to the end of veraison, reaching a peak in almost ripe berries. Whatever the cultivar analysed, the concentration of IAA in berries was highest at the pea-size stage, and then, it decreased strongly under all water regimes assayed (Figure 5b, f). At pea size, ED-treated berries of Graciano had a lower IAA concentration than that of the Control vines, whereas the LD treatment resulted in a significant rise in IAA in almost ripe berries in both cultivars. In general, the concentration of JA reached a maximum in young berries after which it decreased markedly, especially in Graciano (Figure 5c,g). At pea size, the ED treatment provoked an increase of JA in Tempranillo, whereas Graciano showed the opposite trend. Finally, LD stimulated JA production at the beginning of treatment (almost ripe berries) in Tempranillo, but this effect was delayed until harvest in Graciano. The main changes in SA concentration due to water-deficit treatments became evident at the pea-size stage (Figure 5d,h). In the ED treatment, the SA concentration increased in both cultivars, this increase being more marked in Graciano berries, and then it decreased. By contrast, the LD procedure did not produce any significant alteration to SA concentration.
Discussion
Numerous studies have reported that application of a water deficit to grapevines can accelerate (Keller et al. 2008) or delay (Sipiora and Gutiérrez 1998) berry ripening. These inconsistencies obtained in response to seasonal water deficits may be related to differences among studies in the timing of water stress, its severity and duration, and in the cultivar analysed (Esteban et al. 2001, Sivilotti et al. 2005, Girona et al. 2009, Intrigliolo and Castel 2010, Basile et al. 2011, 2012). The present work has attempted to eliminate such variations by subjecting the cultivars Tempranillo and Graciano to a similar water stress level at either pre- or post-veraison, for comparison (Figure 1, Table 1). Therefore, low predawn leaf Ψ in ED-treated plants was indicative of mild water stress (Figure 2a,c) and the lowest value of predawn leaf Ψ recorded in the LD-treated vines after veraison indicated a relatively severe water stress (Deloire et al. 2004, Intrigliolo et al. 2012).
In Tempranillo, despite the important differences in vine water status and vegetative growth, berry growth was almost insensitive to plant water status, but berries showed increased relative skin mass, especially in ED treatment (Table 2). In Graciano, berry growth was significantly reduced by ED treatment but no changes in the relative skin mass were detected. Overall results agree with earlier reports that have shown berry size to be sensitive to water-deficit pre-veraison and much less sensitive post-veraison (Intrigliolo and Castel 2010, Basile et al. 2011, Santesteban et al. 2011, Intrigliolo et al. 2012). The impact of irrigation on all of the must composition parameters studied was highly significant and differed according to the deficit irrigation strategy and grapevine cultivar. In general, irrigation led to a delay in obtaining the desirable sugar concentration, but ED and LD treatments increased sugar accumulation (Table 3). These results are in agreement with those obtained in field-grown Tempranillo (Sipiora and Gutiérrez 1998, Intrigliolo and Castel 2010).
Water-deficit irrigation applied at pre-veraison (ED) or at post-veraison (LD) improved accumulation of phenolic substances in Tempranillo, but no significant effect was detected for Graciano (Table 3). Several studies have reported that the enhanced concentration of phenolic substances in water-stressed plants could be because of berry size reduction, increase in the skin to pulp mass ratio and/or increased biosynthesis of these substances (Esteban et al. 2001, Sivilotti et al. 2005, Castesssllarin et al. 2007b, Intrigliolo and Castel 2010). In the present study, all these effects were strongly dependent on the cultivar studied (Table 2). Consistent with our results, a reduction in the amount of anthocyanins in ED-treated berries has been reported elsewhere by Esteban et al. (2001) in Tempranillo. By contrast, ED and LD in Graciano seemed to be similarly efficient in improving accumulation of anthocyanins, and this might be attributed to higher expression of genes responsible for the anthocyanin biosynthetic pathway during water deficits (Deluc et al. 2009). Our data reinforced the idea that the effect of water deficit on the composition of phenolic substances varies according to the cultivar and timing of water restriction. Some studies have shown that ED improved the accumulation of anthocyanins in different cutlivars and growth conditions (Castellarin et al. 2007b, Basile et al. 2011, Intrigliolo et al. 2012), but the effect of LD treatment was less obvious. Thus, some results have shown that LD can increase (Basile et al. 2011, Santesteban et al. 2011), reduce (Intrigliolo et al. 2012) or cause no change in the concentration of anthocyanins (Sivilotti et al. 2005, Castellarin et al. 2007b, Girona et al. 2009) as occurred in the present study.
Another aspect that should be considered is that highly coloured red grapes do not necessarily produce the most intensely coloured red wines (Holt et al. 2008), and this may be related to the extractability of anthocyanins from grape skins into the must (González-Neves et al. 2004). In our study, the calculated cellular extractability of anthocyanins represents the proportion of non-extracted anthocyanins at pH 3.2 over the maximum possible that was extracted at pH 1 (Figure 3c). A decrease in cellular extractability would indicate that the anthocyanins present in grapes are more easily extracted and that an increase in cellular extractability would indicate more difficulties for the extraction of anthocyanins. The last effect was especially relevant in the case of water-deficit irrigated Graciano and may be indicative of maturation excess (Salazar-Parra et al. 2010). Other effects indicative of maturation excess were decreased malic acid concentration and an increase in pH, as was observed in the present study in plants subjected to ED treatments (Table 3).
Berry amino acids contribute to wine aroma, taste and appearance (Romano et al. 2003). The present study showed that Tempranillo berries have an amino acid concentration higher than that of Graciano berries, suggesting that Tempranillo has greater aroma potential (Figure 4a,c). Moreover, our results clearly indicated specific effects of water deficit on berry nitrogen compounds (Table 4). Similarly, Basile et al. (2012) reported that LD treatment increased amino acid concentration to a lesser extent that the ED treatment in Chardonnay, as occurred in Graciano (Figure 4c). In this cultivar, a higher amino acid concentration was due to the accumulation of tryptophan, threonine and isoleucine (Table 4), which are important because they can affect the aromatic characteristics of wines (Hernández-Orte et al. 2002, Garde-Cerdán et al. 2009). Results suggest that the application of a water-deficit period in Graciano, and especially an ED strategy, can improve its aromatic potential. Although accumulation of amines in berries can reduce wine quality, because amines can be a health risk to sensitive individuals (Ancín-Azpilicueta et al. 2008), a low concentration of amines is always present in berries because putrescine, spermine or spermidine are involved in their development (Gény et al. 1997, Antolín et al. 2008). In Tempranillo, water-deficit treatments did not alter amine content of berries as reported in other studies (Bover-Cid et al. 2006). However, ED and LD treatments induced significant accumulation of amines in Graciano, mainly due to increased putrescine accumulation (Figure 4d), which might arise from the decarboxylation of amino acids as indicated by the lower amino acid concentration in the berries of the LD-treated vines in comparison with that of the ED treatment in Graciano.
The present work was focused on changes in berry hormones, such as ABA, IAA, JA and SA at several stages of growth and ripening (Figure 5). Water stress has been reported to alter mRNA expression patterns associated with hormone metabolism in grape berries that, in turn, could modify the endogenous berry hormonal status (Deluc et al. 2009, Grimplet et al. 2009). Under well-watered conditions, our data showed the highest concentration of ABA during veraison (Figure 5a,e) in agreement with early results showing that ABA accumulates in the grape berry at the beginning of ripening (Antolín et al. 2008, Deluc et al. 2009, Wheeler et al. 2009, Niculcea et al. 2013). Imposition of both types of regulated water deficit, however, altered the pattern of ABA accumulation (Figure 5a,e). In Tempranillo, the prolongation of ABA production over time in LD-treated berries might be related to an improved level of sugars and phenolic substances (Table 3), and anthocyanins (Figure 3a,b). These results are supported by the well-established effect of ABA on the expression of genes involved in anthocyanins production as the initiator of anthocyanins biosynthesis (Koyama et al. 2010, Gagné et al. 2011). In Graciano, this effect was less evident because both ED and LD treatments were similarly effective in improving sugar and anthocyanins production, suggesting that the response of ABA may vary depending upon the cultivar (Sandhu et al. 2011).
In fleshy fruits such as berries, the concentration of IAA declines toward the onset of ripening (Deytieux-Belleau et al. 2007, Böttcher et al. 2010), and the results of the present study corroborated this observation (Figure 5b,f). At the pea-size stage, the most evident changes were detected in the ED treatment of Graciano, in which IAA was reduced, and which suggested that the reduced concentration of IAA at this point of development may have contributed to low berry size (Table 2) (Niculcea et al. 2013). In the LD treatment, an increase in IAA in almost ripe berries could result in a ripening–delaying effect including a delay in the accumulation of sugars and anthocyanins (Böttcher et al. 2010). This does not appear to be the case in our work, however, and it is possible that an increased ABA concentration detected in the LD treatment at this stage may contribute to eliminate a putative ripening–delaying effect of IAA.
Jasmonic acid is present in a variety of plant organs, including fruits. Previous studies have shown that exogenous JA application can contribute to the triggering of ripening in climacteric fruits by influencing ethylene production (Saniewski et al. 1986, Lorenzo et al. 2003). The role, however, of JA in fruit development and/or ripening in non-climacteric fruits such as grapes is poorly understood. The general pattern for JA concentration in berries of both cultivars was similar to that reported previously (Kondo and Fukuda 2001, Niculcea et al. 2013) (Figure 5c,g).The most relevant changes induced by different water irrigation treatments were detected in Graciano. Thus, the ED treatment reduced JA at the pea-size stage suggesting a possible role of this hormone in the reduction of berry size (Peña-Cortés et al. 2005). In contrast, our data also showed that the ED treatment increased the SA concentration at the pea-size stage in both cultivars (Figure 5d,h) being highest in Graciano, which might be related to a role for SA in modulation of the plant response to abiotic stresses such as drought (Klessig and Malamy 1994).
In summary, this study showed that the main effects of water deficits imposed at different berry stages included a reduction in berry size, an increase in phenolic substances such as anthocyanins, and accumulation of amines. In addition, some alteration in the hormonal status of berries at the time of water stress imposition, i.e. pea size and onset of veraison, respectively, might contribute to explain, at least in part, the differential sensitivity of Tempranillo and Graciano to seasonal water stress. To our knowledge, this is the first study that relates interspecific differences in the sensitivity of seasonal water-deficit irrigation to changes in the endogenous hormonal status of berries.
Acknowledgements
The authors thank Faustino Aguirrezábal (Estación de Viticultura y Enología de Navarra (EVENA), Spain) for providing grapevine cuttings, and Amadeo Urdiáin and Mónica Oyárzun for technical assistance during experiments. This project was supported by the Ministerio de Ciencia e Innovación (MCINN BFU2011-26989) of Spain. Maria Niculcea was the recipient of a grant from the Asociación de Amigos de la Universidad de Navarra.
