Advertisement
Research Article

Grape Berry Transpiration Is Determined by Vapor Pressure Deficit, Cuticular Conductance, and Berry Size

Yun Zhang, Markus Keller
Am J Enol Vitic.  2015  66: 454-462  ; DOI: 10.5344/ajev.2015.15038

Abstract

Developmental changes and factors determining grape berry transpiration were investigated in three genetically diverse Vitis cultivars. Transpiration rates were measured on whole clusters, using a custom-designed cluster chamber, and on individual berries by weighing detached berries over time. Results obtained using the two methods were in good agreement. The chamber method verified the assumption that detaching berries does not alter their transpiration and also showed that rachis transpiration was minor compared to whole cluster transpiration. Berry transpiration fluctuated with vapor pressure deficit which was the main determinant of the driving force. The transpiration rate per berry and, to a lesser extent, the cuticular conductance, peaked when berries were red/purple (~13 Brix) and then declined with further ripening. Due to the decline in cuticular conductance during late ripening, the positive linear relationship between berry transpiration rate and surface area weakened after berries ripened. Despite the similar developmental patterns of berry transpiration and cuticular conductance, Concord (V. labruscana) berries consistently had much lower cuticular conductance than Merlot and Syrah (V. vinifera) berries. These results showed that berry transpiration was determined by both external factors (air temperature and relative humidity) and cultivar-specific internal factors (primarily berry surface area and cuticular conductance).

Water comprises from 70% to >90% of the fresh weight of grape berries at maturity and also determines the concentrations of solutes such as sugars, organic acids, minerals, and phenolic compounds (Keller 2015). In grape berries, as in other fleshy fruits, transpiration contributes to fruit weight loss (Rogiers et al. 2004, Clearwater et al. 2012) and affects fruit ripening and solute accumulation (Rebucci et al. 1997, Morandi et al. 2010). Therefore, understanding how developmental changes and other factors influence berry transpiration is important to improve practices to manipulate yield and quality of grapes.

Fruit transpiration is often measured by weighing detached fruits over time (Lang 1990, Rogiers et al. 2004, Morandi et al. 2010). Based on the assumption that detaching does not alter fruit transpiration (Lang 1990), the transpiration rate (E) is calculated as the rate of weight loss over time. However, whether or not cutting off the water supply to the fruit alters its transpiration has not been tested. Poni et al. (2001) challenged this assumption and used a custom-built system to measure whole cluster transpiration in situ, but did not compare their results with the weighing method to validate it. Berry E is usually much lower than that of leaves (Rogiers et al. 2004, Keller 2015), indicating a very low surface conductance (g) to water vapor. The number of stomata on the berry surface is less than one per mm2 (Blanke and Leyhe 1988), and the stomata become partially or completely occluded by epicuticular wax or are transformed into lenticels after fruit set (Blanke and Lenz 1989, Hardie et al. 1996). Therefore, berry transpiration is predominantly cuticular and lacks stomatal regulation (Possingham et al. 1967). As a result, E per unit surface area (A) is strongly affected by environmental factors such as ambient temperature (T) and relative humidity (RH), which determine the vapor pressure deficit (VPD; Gibert et al. 2005, Nobel 2009).

E declines more than five-fold from unripe to ripe berries, both on a per surface area basis and on a per berry basis (Rogiers et al. 2004). A developmental decrease in E has also been reported for kiwifruit berries (Montanaro et al. 2012), and has been attributed to anatomical modifications of the fruit epidermis (Hallett and Sutherland 2005). This indicates that g may change during development. Due to the lack of functional stomata (Blanke and Leyhe 1988), grape berry g consists of the boundary layer conductance (gb) and cuticular conductance (gc). Because gc is one or two orders of magnitude lower than gb (Nobel 2009), gc exerts the main limitation on berry E and thus g ≈ gc (Gibert et al. 2005). Estimates of g or gc have been reported for sweet cherries (Knoche et al. 2000), kiwifruit berries (Clearwater et al. 2012), and tomato fruit (Leide et al. 2007). However, little information is available for grape berry gc.

To better understand developmental changes and other factors determining berry transpiration, we studied berries of three genetically distinct grape cultivars. We first tested the assumption that detaching berries does not alter berry E, which was measured using two different methods and cross-validated. Also, the relationship between VPD and E was studied. Finally, developmental changes and cultivar differences in berry E and gc were investigated.

Materials and Methods

Plant material.

Own-rooted grapevines, Vitis vinifera L. cvs. Merlot and Syrah (planted in 1999) and V. labruscana Bailey cv. Concord (planted in 2003) were used for field measurements and sample collection in 2011 and 2012. These vines were grown with north-south row orientation and 1.8 m × 2.7 m spacing in the experimental vineyard of the Irrigated Agriculture Research and Extension Center in Prosser, WA (lat. 46°17′N; long. 119°44′W; elevation 365 m asl). Mean annual and growing season (April through October) precipitations were 171 mm and 92 mm, respectively, over the study years. All vines were drip-irrigated. Merlot and Syrah were trained to a bilateral permanent cordon with spur pruning and loose vertical shoot-positioning, and Concord was mechanically pruned with no shoot-positioning. No shoot thinning or other canopy management practices were applied during this study.

Two-year-old, own-rooted Syrah and Concord grapevines were grown in white PVC pots containing a mixture of 50% sandy loam, 25% peat moss, and 25% pumice with an addition of 30 g/L dolomite. These vines were grown in an air-conditioned greenhouse (temperature 18 to 30°C; midday photosynthetically active radiation >1000 μmol photons/m2 sec). All pots were irrigated regularly to prevent water stress.

Grape berries are often classified as preveraison, veraison, or postveraison. However, during veraison the berries undergo many physiological processes that do not occur simultaneously or over the same time scale (Coombe 1992), and berry skin color can vary from green to blue on a single cluster. Consequently, berries were grouped into successive developmental stages based on their firmness to the touch and on skin color (Keller and Shrestha 2014, Keller et al. 2015), and were coded as follows: green hard = 1, green soft = 2, blush/pink = 3, red/purple = 4, or blue = 5. After the berries turned blue, two more groups were determined according to total soluble solids (TSS): ripe = 6 (Merlot and Syrah: 20 to 24 Brix; Concord: 18 to 20 Brix) and overripe = 7 (Merlot and Syrah: >24 Brix; Concord: >20 Brix). Strong correlations were obtained between juice solute concentration, measured as TSS using a hand-held refractometer (Atago, Tokyo, Japan), and developmental stages (codes) for each cultivar (r > 0.96, p < 0.001, n = 50).

Model of grape berry transpiration.

Symbols, units of measurement, and parameters used in the grape berry transpiration model are listed in Table 1. According to Fick’s law (Nobel 2009), the rate of berry transpiration (E) can be described as

Embedded Image Eq. 1

where ΔN = Ni – Na and ΔP = Pi – Pa; Ni and Pi are the mole fraction and partial pressure, respectively, of water vapor in the internal air space (or apoplast) of the berry; Na and Pa are the mole fraction and partial pressure, respectively, of water vapor in the ambient air; and P* is the standard atmospheric pressure. From Equation 1, gc can be calculated as

Embedded Image Eq. 2

It is often assumed that the air space inside a plant organ is saturated with pure water, so that Pi equals the partial water vapor pressure of saturated air (Ps; Nobel 2009, Montanaro et al. 2012). However, solutes accumulate in the apoplast of ripening berries (Wada et al. 2009, Keller and Shrestha 2014), which lowers Pi according to Raoult’s law (Nobel 2009). Therefore, Pi was estimated using the water potential of the berry apoplast (Ψi). As a simplification, we assumed that Ψi ≈ Ψπ (the osmotic potential of berry apoplast), and Ψπ was assumed to be equal to the berry osmotic potential (= 0.31 – 0.20 × TSS; Bondada and Keller 2012). Based on the definitions of RH and water potential of water vapor (Nobel 2009),

Embedded Image Eq. 3

where RHi is the relative humidity inside the berry, R is the gas constant, T is the thermodynamic temperature, and Embedded Image is the partial molar volume of liquid water. Combining Equations 2 and 3,

Embedded Image Eq. 4

where RHa is the relative humidity of ambient air.

View this table:
Table 1

Symbols, units of measurement, and parameters used in the text and the equations of the fruit transpiration model.

For comparison, gc was also estimated based on the aforementioned common assumption that Pi = Ps (that is ΔP = VPD).

Measurement of cluster transpiration using a cluster chamber.

A fruit cluster chamber constructed from two halves of a hollow cylinder made of clear acrylic plastic (245 mm in high, 125 mm in diam, and 3 mm wall thickness) was custom-designed to measure whole-cluster transpiration in situ. The chamber opened by pivoting along one side of the two halves, which were connected by screws. A hinge was used on the other side to close the chamber. A 10 mm × 10 mm hole was cut through the top of the chamber. To measure transpiration, a cluster was enclosed in the chamber with its peduncle protruding through the hole. Foam insulation was glued around the hole and along the sides of the half-cylinder to prevent air leakage when the chamber was closed. A temperature sensor and two RH sensors (ADC BioScientific, Herts, UK) were mounted inside the chamber to monitor T and reference and analysis RH. During a preliminary experiment, both chamber and ambient T were tracked using two temperature sensors (Thermochron iButton; Maxim Integrated, San Jose, CA). Over an ambient T range of 22 to 56°C, the chamber T was consistently 1.86°C higher than the ambient T with no sign of overheating inside the chamber (Chamber T = 0.98 × Ambient T + 1.86, r = 0.98, p < 0.001; see also Supplemental Figure 1). A light sensor (ADC BioScientific) was mounted on and perpendicular to the top of the chamber. The chamber was connected with the console of an LCpro+ portable photosynthesis system (ADC BioScientific) through its handle. The rate of air flow to the chamber was 335 ± 5 μmol/sec. Cluster transpiration rate (Ec) was calculated based on the difference in reference and analysis RH and the air flow rate.

To test the assumption that detachment does not alter transpiration, Ec was measured in situ on field-grown Syrah vines continuously for 1 hr. Then the clusters were detached from the vines, kept at the same location in the canopy, and Ec was recorded for another 1 hr. Measurements were recorded automatically every 2 min. Sixteen clusters were tested during different two-hour windows over several days. The Ec of each cluster before and after detaching and the corresponding VPD (calculated from T and RH; Nobel 2009) were compared using a paired t-test (Statistica 7.0; StatSoft, Inc., Tulsa, OK).

In a separate experiment, the contribution of the rachis to Ec was tested. The Ec of 10 clusters of pot-grown Syrah vines was measured using the cluster chamber for 1 hr each in the greenhouse. Then, the rachis (but not the pedicels and receptacles) of each cluster was carefully covered with antitranspirant (2% v/v di-1-p-menthene; Vapor-Gard, Miller, Hanover, PA) using a small paintbrush, and Ec was measured again for 1 hr. A preliminary experiment showed that berry transpiration was reduced by 30% with Vapor-Gard applied on the berry surface (Y. Zhang and M. Keller, unpublished results), which was somewhat less than the previously reported 50% reduction in leaf transpiration (Palliotti et al. 2013). The Ec of each cluster with and without antitranspirant and the corresponding VPD during the measurements were compared using a paired t-test.

Ec of clusters of pot-grown Concord and Syrah vines was measured in the greenhouse under varying VPD. For each cultivar, three clusters were measured at each of three stages: preveraison (all berries were green hard), veraison (berries ranging from blush/pink to red/purple), and postveraison (all berries had turned blue). In addition, three preveraison Syrah clusters were measured outside the greenhouse during a heat wave (maximum ambient T ~45°C). The reason that these stages were used to categorize clusters was the asynchronous development of berries on one cluster during veraison (Coombe 1992). The transpiration rate per berry measured using the cluster chamber ( Embedded Image) was calculated by dividing Ec by the number of berries on each cluster. Relationships between Embedded Image and VPD were evaluated by correlation analysis (Statistica 7.0).

Berry transpiration measurements by weighing detached berries over time.

Berries of Concord, Merlot, and Syrah were randomly sampled in the experimental vineyard from green hard (preveraison) to overripe. They were immediately transferred to the laboratory in a sealed plastic bag inside a cooler. Berries were grouped into developmental stages as before. For each stage, five or ten 10-berry replicates were used to estimate berry transpiration under constant VPD (1.35 kPa) as described in Keller et al. (2015). Berry TSS of an additional five 10-berry replicates was measured at 0 hr. Berry volume was estimated from berry weight and density calculated from tabulated values of TSS and density (density (kg/L) = 0.9984 + 0.0038 × TSS + 0.00001631 × TSS2). Berry A was calculated from berry volume, based on the properties of a sphere. Daily transpiration was expressed both on a per berry basis (Eb) and on a per surface area basis (Eb/A). Next, gc of each cultivar and developmental stage was calculated using Equation 2. The variation in A, Eb, Eb/A, and gc due to cultivar and developmental stage was evaluated by two-way ANOVA (Statistica 7.0). Due to significant interaction among cultivars and developmental stages, one-way ANOVA was then conducted and differences due to developmental stages within each cultivar were evaluated using Fisher’s least significant difference test. The relationship between Eb and A was evaluated by correlation analysis.

An additional set of 100 berries of each cultivar at each of two developmental stages, green hard and blue, were collected intentionally to cover a large range of berry sizes. Their individual Eb and A were estimated as described above, and the relationship between Eb and A was analyzed by correlation analysis.

Cross-validation of berry transpiration rates measured using two methods.

To cross-validate the two methods used to measure berry transpiration, the values obtained using the cluster chamber and by weighing detached berries were compared. First, berry transpiration under the conditions experienced during the chamber measurements was calculated ( Embedded Image) using Equation 1. For this purpose, gc was obtained using the weighing method, and ΔP was calculated from T and RH recorded by the cluster chamber using Equation 4. Then, the relationship between Embedded Image and Embedded Image was evaluated using correlation analysis. Second, daily berry transpiration rate per surface area under laboratory conditions was calculated (Eb/A′) using the regression equations between Embedded Image and VPD obtained from the cluster chamber experiments. Then Eb/A′ was compared with Eb/A at three stages (preveraison, veraison, and postveraison). No statistical analysis could be conducted for this comparison because the calculated Eb/A′ could not be replicated. Therefore, the comparison between Eb/A′ and Eb/A was only to test whether the calculated values were in the same range as the measured ones.

Results

Detaching the clusters from field-grown vines did not change Ec significantly when compared with Ec measured before detaching (32 ± 11 and 35 ± 11 μmol H2O/sec, before and after detaching, respectively, n = 16, p = 0.54). No difference in Ec was found before and after the rachis of pot-grown vines was covered with antitranspirant (5.3 ± 1.3 and 5.4 ± 1.2 μmol H2O/sec, with or without antitranspirant, respectively, n = 10, p = 0.45). Therefore, cluster transpiration was dominated by berry transpiration. Due to differences in VPD (2.7 kPa in the field and 1.9 kPa in the greenhouse) and cluster size (≥150 berries per cluster on field-grown vines but only 30 to 90 berries per cluster on pot-grown vines), the absolute values of Ec between these two experiments were rather different. Nevertheless, this difference was irrelevant to the abovementioned conclusion.

Diurnal changes in Embedded Image closely followed fluctuations in VPD (Figure 1). Daily berry E varied between 14 and 20 mg H2O/day during this experiment. Furthermore, Embedded Image exhibited a close linear relationship with VPD from preveraison to postveraison (Figure 2), even when VPD went up to 8 kPa during a heat wave (inset in Figure 2A). Under the same VPD, Embedded Image was slightly higher at veraison than at either pre- or postveraison (Figure 2), which implies a developmental change in berry gc.

Figure 1
Figure 1

Diurnal changes in grape berry transpiration rate ( Embedded Image) and vapor pressure deficit (VPD) measured over five days on a preveraison cluster of a pot-grown Syrah vine using a cluster chamber. Numbers underneath the peaks indicate the integrated daily berry transpiration (mg H2O/day).

Figure 2
Figure 2

Correlation between vapor pressure deficit (VPD) and grape berry transpiration rate ( Embedded Image) measured on potted vines at preveraison (all berries were green hard), veraison (berries varied from blush/pink to red/purple), and postveraison (all berries were blue) using a cluster chamber. (A) Syrah: preveraison: y = 40x − 3; veraison: y = 47x + 1; postveraison: y = 41x − 6; (B) Concord: preveraison: y = 25x − 3; veraison: y = 38x − 10; postveraison: y = 29x − 6. All r ≥ 0.97, p < 0.001, n ≥ 650. Inset in (A) shows a similar relationship between VPD and ( Embedded Image) of preveraison Syrah clusters at much higher VPD (y = 41x −11, r = 0.99, p < 0.001, n = 564).

Average daily Eb for all three cultivars, measured by weighing detached berries, was lowest at the green hard stage (20.8 to 24.9 mg H2O/day) and peaked at the red/purple stage (p < 0.001), constituting a 1.7-, 1.7-, and 2.4-fold increase for Concord, Merlot, and Syrah, respectively (Figure 3A). Eb declined with further ripening (p < 0.001; 1.2-, 1.5-, and 1.6-fold decrease for Concord, Merlot, and Syrah, respectively; Figure 3A). Although Concord berries were much larger than Merlot and Syrah berries (Figure 3C), the Eb of Concord was either less than or in between that of Merlot and Syrah. Developmental changes in Eb/A exhibited a similar pattern as Eb, peaking at the red/purple stage for these three cultivars (p < 0.001; Figure 3B). No difference in Eb/A was found between Merlot and Syrah (p = 0.92), except at the ripe and overripe stages (p < 0.001). However, the average Eb/A of Concord was 40% lower than that of Merlot and Syrah (p < 0.001). Since the measurements were taken in the same environment for all cultivars, the differences in Eb/A imply genetic differences in berry gc.

Figure 3
Figure 3

Grape berry (A) transpiration rate per berry (Eb), (B) transpiration rate per surface area (Eb/A), and (C) surface area of Concord, Merlot, and Syrah at successive developmental stages (1 = green hard; 2 = green soft; 3 = blush/pink; 4 = red/purple; 5 = blue; 6 = ripe; 7 = overripe). The color of the circle above the x-axis indicates the color of the berry skin at each stage. Transpiration was measured by weighing detached berries over 48 hr in a controlled environment. Data are means ± SE where SE > symbol size (n ≥ 5, 10 berries per replicate).

When measured under identical conditions, Eb increased linearly as berry A increased from the green hard to the blue stage (TSS ≤ 20 Brix; Figure 4A). However, the relationship between A and Eb became weaker when berries with TSS >20 Brix (ripe and overripe stages) were included in the analysis (Figure 4B). This implies that, irrespective of cultivar, a decrease in gc was responsible for the lower Eb at advanced ripening stages. The relationship between Eb and A was further tested in 100 additional berries of each cultivar at the green hard and blue stages. Again, Eb exhibited positive linear relationships with A, which ranged from 225 to 584 mm2 for green hard berries, and from 296 to 1059 mm2 for blue berries (Figure 5). These linear relationships under the same VPD at a specific developmental stage also implied constant gc at that stage. At both stages, the slopes of the linear regressions for Merlot and Syrah were similar, while for Concord, Eb increased less with increasing A, again indicating lower berry gc for Concord than Merlot and Syrah.

Figure 4
Figure 4

Relationship between grape berry surface area and berry transpiration rate (Eb). (A) Berries from green hard to blue stages (total soluble solids (TSS) ≤ 20 Brix, n = 25 per cultivar). (B) Ripe and overripe berries (20 Brix ≤ TSS ≤ 27 Brix) added to the berries presented in (A) (n = 35 per cultivar). Transpiration was measured by weighing detached berries over 48 hr in a controlled environment.

Figure 5
Figure 5

Correlation between grape berry surface area and berry transpiration rate (Eb) of Concord, Merlot, and Syrah at green hard (A) and blue (B) stages. (A) Concord: y = 0.05x + 2, r = 0.61; Merlot: y = 0.15x − 20, r = 0.80; Syrah: y = 0.15x − 18, r = 0.89. (B) Concord: y = 0.04x + 10, r = 0.81; Merlot: y = 0.08x − 2, r = 0.92; Syrah: y = 0.10x − 7, r = 0.96. All p < 0.001, n = 100. Transpiration was measured by weighing detached berries over 48 hr in a controlled environment.

Consistent with this notion, gc calculated using Equation 2 was much higher overall for Merlot and Syrah than for Concord (p < 0.001; Table 2). For Concord berries, gc was highest at the red/purple and blue stages (p < 0.001). After the berries turned ripe, gc declined to its lowest level (p < 0.001). Similar to Concord, gc of Merlot and Syrah berries was highest at the red/purple stage (p < 0.001). Different from Concord, gc started decreasing as Merlot and Syrah berries progressed from red/purple to blue (p < 0.001). Although Ψi decreased from −0.8 MPa to −4.8 MPa during ripening, this only caused a slight decrease in RHi (from 0.99 to 0.97) and ΔP (from 1.28 kPa to 1.22 kPa). Consequently, under the conditions of this study, the assumption Pi = Ps led to an overestimation of Pi by 0.7 (preveraison) to 3.5% (overripe) and an underestimation of gc by 1.4 to 7.3%.

View this table:
Table 2

Cuticular conductance (gc) of Concord, Merlot, and Syrah grape berries at varying developmental stages. The gc was calculated as the rate of berry transpiration per surface area divided by the gradient of the mole fraction of water vapor between the ambient air and the berry apoplast (Equation 2). Data are presented as means ± SE (n = 5; 10 berries per replicate). Different upper case letters indicate significant difference among genotypes within each stage; different lower case letters indicate significant difference among stages within each genotype (p < 0.05, Fisher’s least significant difference test).

The results for berry transpiration obtained using two different methods, the cluster chamber and weighing detached berries, were cross-validated. The Embedded Image (calculated using gc from the weighing method) had a close linear relationship with Embedded Image measured for both Syrah and Concord (Figure 6). The relationship for Syrah was close to 1:1, but for Concord, the calculated Embedded Image underestimated the measured Embedded Image. The Eb/A′ calculated using the regression equations obtained from the chamber measurements (see Figure 2) was in the same range as Eb/A measured in the weighing experiments (Table 3). The largest variation was at veraison (12.2 and 16.7 g H2O/m2 day for Syrah and Concord, respectively), which was probably due to the asynchrony of berry ripening within one cluster at veraison.

Figure 6
Figure 6

Correlation between grape berry transpiration rates measured ( Embedded Image) using a cluster chamber from preveraison to postveraison and estimated ( Embedded Image) using temperature and relative humidity recorded by the chamber and cuticular conductance. Dotted line indicates 1:1 ratio. (A) Syrah: y = 8 + 0.83x; (B) Concord: y = 10 + 0.52x. For both, r ≥ 0.96, p < 0.001, and n > 1500.

View this table:
Table 3

Comparison of two methods to estimate grape berry transpiration. Weighing method: transpiration per surface area was measured by weighing detached berries (Eb/A, means ± SE) at the green hard stage (preveraison); the blush/pink and red/purple stages (veraison); and the blue, ripe, and overripe stages (postveraison). Cluster chamber method: transpiration per surface area was calculated (Eb/A′) using the regression equations for berry transpiration and vapor pressure deficit from the cluster chamber experiment at preveraison, veraison, and postveraison (see Figure 2).

Discussion

The presented results demonstrate that grape berry transpiration is mainly driven by VPD (Figures 1 and 2), depends on berry surface area (that is berry size; Figures 4 and 5), and is modulated by gc (Table 2). Although berries transpire at much lower rates than do leaves (Keller et al. 2015), unlike leaves, ripening berries are unable to control their E via stomatal closure (Possingham et al. 1967, Blanke and Leyhe 1988). Therefore, berry E closely traces changes in VPD. Nevertheless, under the same VPD, berry E differs due to both berry size and gc, both of which vary by cultivar and developmental stage. All three cultivars tested here shared the same developmental pattern of berry E and gc (Figure 3A and B, Table 2). The transient increase in Eb observed at the beginning of ripening was mostly attributable to berry expansion during the second growth phase (37, 45, and 82% increase in berry surface area of Concord, Merlot, and Syrah, respectively; Figure 3C) and, to a lesser extent, an increase in gc (18, 18, and 31% increase in Concord, Merlot, and Syrah, respectively; Table 2). During late ripening, the decline in gc (Table 2) resulted in a decline in Eb despite increasing or stable berry surface area (Figures 3C and 4B). Thus, the positive linear relationship between Eb and A under the same VPD was sustained until the blue stage (Figure 4A), and then disappeared once the berries exceeded 20 Brix (ripe and overripe stages; Figure 4B). This suggests that, under constant VPD, the increase in berry A dominated changes in Eb during early ripening, but the decrease in gc became dominant during late ripening. The three cultivars varied significantly in the absolute values of gc (Table 2), which led to Eb/A in Merlot and Syrah being nearly twice that of Concord (Figure 3B).

Fruit transpiration rate is often estimated by measuring weight loss in detached fruits (Rogiers et al. 2004, Morandi et al. 2010, Becker and Knoche 2011). Our results support the assumption that detaching does not alter fruit transpiration, at least in the short term. Using the cluster chamber, no difference in Ec was found before and after detaching. Previously, we reported that weight loss from detached berries remained constant over three successive 24-hr periods (Keller et al. 2015). Also, rachis transpiration was insignificant for whole-cluster transpiration. This may be partly due to the small surface area of the rachis (~4%) relative to the whole cluster surface area (Poni et al. 2001). This conclusion is similar to the finding that the contribution of the pedicel/receptacle to overall berry transpiration was negligible (Becker and Knoche 2011). Furthermore, the rates of berry transpiration obtained using the cluster chamber and the weighing method agreed with each other (Table 3 and Figure 6).

The strong positive correlation between Embedded Image and VPD (Figures 1 and 2) indicates that VPD (thus T and RH) is the dominant determinant of the driving force (ΔP) for berry transpiration. Unlike leaves, which may close their stomata to reduce transpiration, grape berries do not regulate their transpiration (Possingham et al. 1967), hence the linear relationship between Embedded Image and VPD even up to 8 kPa (inset of Figure 2). This suggests that berries may be vulnerable to dehydration under high VPD. Although Eb was much lower than the leaf transpiration rate (2 to 14 mmol H2O/m2 sec; Keller 2015), it accounted for a daily loss of 2 to 5% of berry weight (23 to 54 mg H2O/day; Figure 3A) in a controlled laboratory environment (VPD = 1.35 kPa). Daily mean VPD in the field during ripening may be higher (up to 2.7 kPa under the climatic conditions of this study; http://weather.wsu.edu). Based on the weather data during ripening of the two experimental years, Eb integrated over 24 hr varied between 3 and 100 mg H2O/day. This is equivalent to 0.2 to 6% of berry fresh weight and is largely due to daily variation in VPD.

Our method of categorizing and measuring berries according to their firmness to the touch and skin color permitted insight into the details of the developmental changes in E during veraison. Using this sampling approach, we observed a two-phase pattern of developmental change in berry E that was consistent among the three cultivars (Figure 3A). This contradicts the previous finding of a steady decline in berry E from preveraison to postveraison (Rogiers et al. 2004). These authors may have missed the possible changes during veraison (e.g., TSS changed from 8 to 20 Brix from the green soft to the blue stage; Supplemental Figure 2) because no sample was taken during veraison (see Table 1 in Rogiers et al. 2004). Close examination of Figure 3B in Becker and Knoche (2011) also suggests that there might be a slight increase in berry E from 60 to 70 days after full bloom in Riesling.

Overall, grape berry gc (2 to 4 mmol H2O/m2 sec; Table 2) is comparable to the gc of grape leaves (~5 mmol H2O/m2 sec; Boyer et al. 1997), kiwifruit berries (~4 mmol H2O/m2 sec; Clearwater et al. 2012), and tomato fruit (0.4 to 6 mmol H2O/m2 sec; Leide et al. 2007). Yet the developmental pattern of grape berry gc differed from that of kiwifruit berries and tomato fruit (Leide et al. 2007, Montanaro et al. 2012). Instead of declining during development, grape berry gc of three cultivars consistently increased from the green hard to the red/purple stage, then decreased with further ripening (Table 2). This developmental pattern in gc could explain why, on a per surface area basis, berry E at the same VPD was higher during veraison than preveraison or postveraison (Figures 2 and 3B). Also, the decline in gc during late ripening most likely contributed to the poor correlation between Eb and berry A (Figure 4B). Since the amount of cuticular wax per surface area does not increase during berry ripening (Grncarevic and Radler 1971, Rogiers et al. 2004), the decline in gc after berries turned red/purple (Table 2) may be due to modifications in the chemical composition of the wax (Possingham et al. 1967, Riederer and Schreiber 2001, Leide et al. 2007). Comparing berry E with different fractions of the cuticular wax removed to that of control berries, Grncarevic and Radler (1971) concluded that soft wax, which consists of long-chain aliphatic compounds (e.g., alcohols, aldehydes, esters, free acids, and hydrocarbons; Radler 1965a), forms the main barrier to water evaporation. The significance of soft wax in fruit transpiration has also been demonstrated using the tomato mutant lecer6 (Leide et al. 2007). This mutant is defective in a β-ketoacyl-coenzyme A synthase, an enzyme involved in very-long-chain fatty acid elongation (Vogg et al. 2004). The cuticle of lecer6 fruit exhibited a decrease in very-long-chain alkanes and a distinct increase in water permeability compared to wild-type fruit (Leide et al. 2007). The amount of soft wax in grape berry cuticles increases from 40 days after full bloom (around veraison) to maturity (Yamamura and Naito 1983). This reported increase in soft wax coincides with the observed decline in Eb/A and gc later during ripening (Figure 3B and Table 2). However, we currently do not have an explanation for the transient increase in gc during early ripening (Table 2). In addition to these developmental changes, the amounts of soft and total wax on V. labruscana berries are more than twice those on V. vinifera berries (Grncarevic and Radler 1971, Yamamura and Naito 1983). Furthermore, the soft wax of V. labruscana berries may be much richer in hydrophobic hydrocarbons than that of V. vinifera berries (Radler 1965b). These findings are in agreement with the significantly lower Eb/A and gc of Concord (V. labruscana) compared with Merlot or Syrah (V. vinifera) observed in this study (Figure 3B).

The relationships among internal and external factors that influence berry transpiration are presented in a flow chart (Figure 7). Berry surface area, Ψi, and gc are internal factors under both genetic and environmental control: surface area increases with berry growth (Figure 3C), Ψi decreases with berry sugar accumulation, while gc increases then decreases during berry ripening (Table 2). Because the effect of Ψi on ΔP is rather small, ambient T and RH (thus VPD) are the dominant factors determining the driving force (ΔP) for transpiration (Figures 1 and 2). Under constant VPD, Eb correlates positively with berry A (Figures 4A and 5; Becker and Knoche 2011); however, this relationship no longer holds once berries exceed 20 Brix due to the decline in gc (Figure 4B and Table 2). Consequently, berry E changes diurnally due to fluctuations in VPD and developmentally due to berry growth and changes in gc. Finally, despite a similar developmental pattern, cultivars differ in their berry E due to variations in berry size and gc.

Figure 7
Figure 7

Internal (shaded rectangles) and external (open rectangles) factors that influence the rate of grape berry transpiration (Eb). Berry surface area (A), apoplast water potential (Ψi), and cuticular conductance (gc) change during berry development. Ambient air temperature (T) and relative humidity (RHa), i.e., vapor pressure deficit, are the main factors determining the driving force (ΔP) for berry transpiration, while the influence of Ψi is minor. Consequently, Eb is determined by ΔP, gc and A.

Conclusion

Berry transpiration of three grape cultivars was measured using two methods and the results were in good agreement. Variations in VPD, berry size, and cuticular conductance determined daily and developmental changes and cultivar differences in berry transpiration rate. Being predominantly cuticular, berry transpiration was strongly dependent on VPD, which resulted in pronounced diurnal fluctuations. Developmentally, berry transpiration (both per berry and, to a lesser extent, per surface area) peaked when berries were red/purple. On a per berry basis, transpiration correlated strongly with berry surface area from the green hard to blue stages. This relationship did not hold after berries exceeded 20 Brix due to a decline in cuticular conductance. This developmental pattern in berry transpiration was evident in all three cultivars. Although Concord had much lower cuticular conductance than Merlot and Syrah, transpiration per berry was similar among the three cultivars due to the greater size of Concord berries.

Acknowledgments

Acknowledgments: This study was supported by funds from the Chateau Ste. Michelle Distinguished Professorship, the Department of Horticulture at Washington State University, and the Rhone Rangers. The authors thank John Ferguson and Giulio Carmassi for technical assistance.

Footnotes

  • Supplemental data is freely available with the online version of this article at www.ajevonline.org.

  • Received May 2015.
  • Revision received July 2015.
  • Accepted July 2015.
  • Published online October 2015

This is an open access article distributed under the CC BY license https://creativecommons.org/licenses/by/4.0/.

Literature Cited

  1. ↵
    1. Becker T. and
    2. Knoche M.
    . 2011. Water movement through the surfaces of the grape berry and its stem. Am. J. Enol. Vitic. 62:340–350.
    OpenUrlAbstract/FREE Full Text
  2. ↵
    1. Blanke M.M. and
    2. Leyhe A.
    . 1988. Stomatal and cuticular transpiration of the cap and berry of grape. J. Plant Physiol. 132:250–253.
    OpenUrl
  3. ↵
    1. Blanke M.M. and
    2. Lenz F.
    . 1989. Fruit photosynthesis. Plant Cell Environ. 12:31–46.
    OpenUrlCrossRef
  4. ↵
    1. Bondada B.R. and
    2. Keller M.
    . 2012. Morphoanatomical symptomatology and osmotic behavior of grape berry shrivel. J. Am. Soc. Hortic. Sci. 137:20–30.
    OpenUrl
  5. ↵
    1. Boyer J.S.,
    2. Wong S.C. and
    3. Farquhar G.D.
    . 1997. CO2 and water vapor exchange across leaf cuticle (epidermis) at various water potentials. Plant Physiol. 114:185–191.
    OpenUrlAbstract
  6. ↵
    1. Clearwater M.J.,
    2. Luo Z.,
    3. Ong S.E.C.,
    4. Blattmann P. and
    5. Thorp T.G.
    . 2012. Vascular functioning and the water balance of ripening kiwifruit (Actinidia chinensis) berries. J. Exp. Bot. 63:1835–1847.
    OpenUrlCrossRefPubMed
  7. ↵
    1. Coombe B.G.
    1992. Research on development and ripening of the grape berry. Am. J. Enol. Vitic. 43:101–110.
    OpenUrlAbstract/FREE Full Text
  8. ↵
    1. Gibert C.,
    2. Lescourret F.,
    3. Génard M.,
    4. Vercambre G. and
    5. Pérez Pastor A.
    . 2005. Modelling the effect of fruit growth on surface conductance to water vapour diffusion. Ann. Bot. 95:673–683.
    OpenUrlCrossRefPubMed
  9. ↵
    1. Grncarevic M. and
    2. Radler F.
    . 1971. A review of the surface lipids of grapes and their importance in the drying process. Am. J. Enol. Vitic. 22:80–86.
    OpenUrlAbstract/FREE Full Text
  10. ↵
    1. Hallett I.C. and
    2. Sutherland P.W.
    . 2005. Structure and development of kiwifruit skins. Int. J. Plant Sci. 166:693–704.
    OpenUrlCrossRef
  11. ↵
    1. Hardie W.J.,
    2. O’Brien T.P. and
    3. Jaudzems V.G.
    . 1996. Morphology, anatomy and development of the pericarp after anthesis in grape, Vitis vinifera L. Aust. J. Grape Wine Res. 2:97–142.
    OpenUrlCrossRef
  12. ↵
    1. Keller M.
    2015. The Science of Grapevines: Anatomy and Physiology. 2nd edition. Elsevier Academic Press, Burlington, MA.
  13. ↵
    1. Keller M. and
    2. Shrestha P.M.
    . 2014. Solute accumulation differs in the vacuoles and apoplast of ripening grape berries. Planta 239:633–642.
    OpenUrlPubMed
  14. ↵
    1. Keller M.,
    2. Zhang Y.,
    3. Shrestha P.M.,
    4. Biondi M. and
    5. Bondada B.R.
    . 2015. Sugar demand of ripening grape berries leads to recycling of surplus phloem water via the xylem. Plant Cell Environ. 38:1048–1059.
    OpenUrlCrossRef
  15. ↵
    1. Knoche M.,
    2. Peschel S.,
    3. Hinz M. and
    4. Bukovac M.J.
    . 2000. Studies on water transport through the sweet cherry fruit surface: Characterizing conductance of the cuticular membrane using pericarp segments. Planta 212:127–135.
    OpenUrlCrossRefPubMed
  16. ↵
    1. Lang A.
    1990. Xylem, phloem and transpiration flows in developing apple fruits. J. Exp. Bot. 41:645–651.
    OpenUrlCrossRef
  17. ↵
    1. Leide J.,
    2. Hildebrandt U.,
    3. Reussing K.,
    4. Riederer M. and
    5. Vogg G.
    . 2007. The developmental pattern of tomato fruit wax accumulation and its impact on cuticular transpiration barrier properties: Effects of a deficiency in a β-ketoacyl-coenzyme A synthase (LeCER6). Plant Physiol. 144:1667–1679.
    OpenUrlAbstract/FREE Full Text
  18. ↵
    1. Montanaro G.,
    2. Dichio B.,
    3. Xiloyannis C. and
    4. Lang A.
    . 2012. Fruit transpiration in kiwifruit: Environmental drivers and predictive model. AoB PLANTS doi:10.1093/aobpla/pls036.
    OpenUrlCrossRefPubMed
  19. ↵
    1. Morandi B.,
    2. Manfrini L.,
    3. Losciale P.,
    4. Zibordi M. and
    5. Corelli-Grappadelli L.
    . 2010. The positive effect of skin transpiration in peach fruit growth. J. Plant Physiol. 167:1033–1037.
    OpenUrlPubMed
  20. ↵
    1. Nobel P.S.
    2009. Physiochemical and Environmental Plant Physiology. 4th edition. Elsevier Academic Press, Burlington, MA.
  21. ↵
    1. Palliotti A.,
    2. Panara F.,
    3. Famiani F.,
    4. Sabbatini P.,
    5. Howell G.S.,
    6. Silvestroni O. and
    7. Poni S.
    . 2013. Postveraison application of antitranspirant di-1-p-menthene to control sugar accumulation in Sangiovese grapevines. Am. J. Enol. Vitic. 64:378–385.
    OpenUrlAbstract/FREE Full Text
  22. ↵
    1. Poni S.,
    2. Intrieri C.,
    3. Rebucci B.,
    4. Magnanini E. and
    5. Filippetti I.
    . 2001. A custom-built simple system for conditioning and measurement of in situ whole-cluster transpiration. Vitis 40:55–58.
    OpenUrl
  23. ↵
    1. Possingham J.V.,
    2. Chambers T.C.,
    3. Radler F. and
    4. Grncarevic M.
    . 1967. Cuticular transpiration and wax structure and composition of leaves and fruit of Vitis vinifera. Aust. J. Bio. Sci. 20:1149–1153.
    OpenUrl
  24. ↵
    1. Radler F.
    1965a. Reduction of loss of moisture by the cuticle wax components of grapes. Nature 207:1002–1003.
    OpenUrl
  25. ↵
    1. Radler F.
    1965b. The main constituents of the surface waxes of varieties and species of the genus Vitis. Am. J. Enol. Vitic. 16:159–167.
    OpenUrlAbstract/FREE Full Text
  26. ↵
    1. Rebucci B.,
    2. Poni S.,
    3. Intrieri C.,
    4. Magnanini E. and
    5. Lakso A.
    . 1997. Effects of manipulated grape berry transpiration on post-veraison sugar accumulation. Aust. J. Grape Wine Res. 3:57–65.
    OpenUrl
  27. ↵
    1. Riederer M. and
    2. Schreiber L.
    . 2001. Protecting against water loss: Analysis of the barrier properties of plant cuticles. J. Exp. Bot. 52:2023–2032.
    OpenUrlCrossRefPubMed
  28. ↵
    1. Rogiers S.Y.,
    2. Hatfield J.M.,
    3. Jaudzems V.G.,
    4. White R.G. and
    5. Keller M.
    . 2004. Grape berry cv. Shiraz epicuticular wax and transpiration during ripening and preharvest weight loss. Am. J. Enol. Vitic. 55:121–127.
    OpenUrlAbstract/FREE Full Text
  29. ↵
    1. Vogg G.,
    2. Fischer S.,
    3. Leide J.,
    4. Emmanuel E.,
    5. Jetter R.,
    6. Levy A.A. and
    7. Riederer M.
    . 2004. Tomato fruit cuticular waxes and their effects on transpiration barrier properties: Functional characterization of a mutant deficient in a very-long-chain fatty acid β-ketoacyl-CoA synthase. J. Exp. Bot. 55:1401–1410.
    OpenUrlCrossRefPubMed
  30. ↵
    1. Wada H.,
    2. Matthews M.A. and
    3. Shackel K.A.
    . 2009. Seasonal pattern of apoplastic solute accumulation and loss of cell turgor during ripening of Vitis vinifera fruit under field conditions. J. Exp. Bot. 60:1773–1781.
    OpenUrlCrossRefPubMed
  31. ↵
    1. Yamamura H. and
    2. Naito R.
    . 1983. The surface wax of several grapes in Japan. J. Jpn. Soc. Hortic. Sci. 52:266–272.
    OpenUrl
Back to top
Print
View full PDF
Email Article
Citation Tools

Open Access
Grape Berry Transpiration Is Determined by Vapor Pressure Deficit, Cuticular Conductance, and Berry Size
Yun Zhang, Markus Keller
Am J Enol Vitic.  2015  66: 454-462  ; DOI: 10.5344/ajev.2015.15038
del.icio.us logo Twitter logo Facebook logo Mendeley logo
Save to my folders

Jump to section