Abstract
Background and goals Sugar import via the phloem increases the sugar content of ripening grape berries while the berries expand. After phloem import ceases, concentration of these sugars may occur as the berries shrink due to dehydration. This study sought to separate the phases of active and passive increases in berry sugar concentration, and to determine the transition point between these phases.
Methods and key findings Cabernet Sauvignon, Merlot, and Syrah berries were sampled in a vineyard in warm and arid southeastern Washington in 2017 and 2018. Grape berry weight, total soluble solids (TSS), titratable acidity, pH, malate, and total hexoses (glucose + fructose) were determined repeatedly from the late lag phase of berry development through late ripening. The berries expanded for 2 to 3 wk after veraison and actively accumulated sugar while breaking down malate for at least another week after berry growth ceased. Depending on the cultivar, the TSS concentration was ~23 Brix when sugar import stopped, but continued to increase passively as the berries gradually lost 6 to 15% of their weight while their pH increased.
Conclusions and significance This study showed that in the three winegrape cultivars examined, berry sugar concentrations above 23 to 24 Brix are reached through berry dehydration, which is necessarily associated with yield loss. These results have direct practical implications since TSS thresholds and weight loss during extended ripening can be readily estimated by repeated and systematic berry sampling.
Introduction
Ripening grape berries accumulate sugar to very high concentrations while simultaneously increasing in size, at least for a while. At a total soluble solids (TSS) concentration of 24 Brix, berries contain ~120 g/L each of the two hexoses, glucose and fructose, and have a total solute concentration of ~1.4 M (Hrazdina et al. 1984, Keller 2020). Sugar accumulation relies mostly on sucrose import via the phloem and is enabled by the symplastic isolation of the berry mesocarp cells from the release phloem at the onset of ripening (Zhang et al. 2006). The sucrose is unloaded from the phloem to the berry apoplast (i.e., cell walls and intercellular spaces); partly or fully hydrolyzed by invertases located in the cell walls, cytoplasm, and vacuoles; and transported across the cellular and vacuolar membranes for hexose storage in the vacuoles (Sarry et al. 2004, Keller 2020). Phloem inflow increases greatly after berry softening at the onset of ripening to support the berries’ massive sugar demand (Terrier et al. 2005, Keller et al. 2015, Zhang and Keller 2017, Hernández-Montes et al. 2021). The rates of sugar import and active accumulation then gradually decline and eventually cease, but the upper bounds and underlying causes for the arrest in sugar storage remain unclear.
Berry sugar accumulation is accompanied by malate catabolism as well as by water inflow via the phloem. As sugars are allocated to storage in the mesocarp vacuoles, the berries switch to respiration of previously stored malate to help meet their energy requirements (Hrazdina et al. 1984, Sarry et al. 2004, Shahood et al. 2020). A small fraction of malate may also be converted back to glucose in the process of gluconeogenesis (Famiani et al. 2016). Some of the phloem-derived water is used for renewed berry growth, but most is discharged by outflow via the xylem (i.e., xylem backflow) and evaporation from the berry surface (Greenspan et al. 1994, Keller et al. 2015). The rate of backflow depends on the sugar concentration of the phloem sap and rises and falls in concert with the temporal change in phloem inflow throughout ripening (Zhang and Keller 2017). Surface evaporation (i.e., berry transpiration) declines following a brief rise associated with berry enlargement during early ripening, but it is essentially a linear function of vapor pressure deficit (VPD) at any stage of ripening since grape berries lack stomatal control (Rogiers et al. 2004, Zhang and Keller 2015).
Prior work suggested that sugar accumulation, and hence phloem import, might continue for some time, albeit at a slow rate, even as grape berries begin to shrink during the late ripening phase (Rogiers et al. 2006, Keller et al. 2015). Based on results from single-berry sampling, however, it has been argued that berry growth and sugar accumulation stop simultaneously as soon as a berry reaches its maximum size (Shahood et al. 2020). Nevertheless, the same team also found that expression of some berry sugar transporter genes dropped 80 to 90% (rather than 100%) once the berries stopped expanding (Savoi et al. 2021), which is consistent with the 86% reduction in sugar accumulation reported by Keller et al. (2015). Regardless, once sugar import has stopped, the amount of sugar per berry (i.e., sugar content) cannot increase further, but the sugar concentration may continue to rise if a berry shrinks due to water loss.
Up until the late 20th century, acceptable grape maturity for winemaking was often subjectively defined as a TSS of 18 to 20 Brix (e.g., Huglin and Schneider 1998). Any time interval that grapes were left on the vine beyond that soft threshold would have been considered “extended ripening.” Today, many winegrapes are routinely harvested at much higher TSS concentrations, often exceeding 25 Brix. For grapegrowers, late-season berry shrinkage due to dehydration associated with extended ripening equates to a loss of yield (La Rosa and Nielson 1956). Additionally, the longer that grapes remain on the vine, the risk of crop loss resulting from diseases and fall frost increases. It would therefore be desirable to establish the ripening point at which active sugar accumulation associated with phloem import ceases and the sugar concentration starts to increase passively due to continued water loss. An earlier study evaluated the sequence of events that occur during the onset of berry ripening (Hernández-Montes et al. 2021). Here, we focus mostly on the late stages of ripening above a berry TSS concentration of 18 to 20 Brix. Hypothesizing that berry growth and sugar accumulation cease simultaneously when grape berries reach their maximum weight (Shahood et al. 2020), our main goal was to identify the timing and TSS threshold at the transition from active sugar accumulation to passive increase in sugar concentration as the berries of three major red wine cultivars begin to shrink during late ripening.
Materials and Methods
Vineyard site and management
We used own-rooted Vitis vinifera L. cvs. Cabernet Sauvignon (planted in 1994), Merlot (planted in 1994), and Syrah (planted in 1998), growing in adjacent 1-ha blocks within a 48.5-ha vineyard located in the Red Mountain American Viticultural Area (AVA) in southeastern Washington (46°16’N; 119°27’W; 211 m asl). The soils in all three blocks are Scooteney silt loam and Warden silt loam (https://websoilsurvey.sc.egov.usda.gov). All vines are planted at 2.7 m between rows and 1.2 m within rows oriented between 0° and 12.5° off true north down a <5% slope, trained to bilateral cordons, and spur pruned to 28 to 30 buds per vine. About one-third of the shoots are vertically positioned, with the remainder positioned evenly to the east and west using foliage wires to allow only dappled sunlight to reach the fruit clusters. Clusters are thinned prior to veraison to achieve target yields of 7, 9.5, and 10 t/ha in Cabernet Sauvignon, Merlot, and Syrah, respectively. Pest and disease control follows integrated pest management best practices backed by research and extension recommendations. The vineyard floor between rows supports a permanent cover crop of crested wheatgrass, and the undervine area is managed using a combination of herbicide application in spring and mechanical cultivation after bloom. Mineral nutrients are applied by fertigation after budbreak and prior to bloom for a total of 19 kg N, 5 kg P, and 3 kg K/ha. Micronutrients are applied as needed, but none were applied in 2017 or 2018.
Soil moisture is monitored at 15-min intervals using Capacitance probes (AquaCheck) at 20-cm intervals to a depth of 1.2 m, and midday stem water potential (Ψs) is monitored using a pressure chamber (model 615, PMS Instrument Company) on 10 leaves per block that have been bagged for 1 hr between 1230 and 1300 hr three to four times per week. Drip irrigation is applied using 1.6-L/hr emitters spaced 1.2 m apart. The vineyard is irrigated to soil saturation down to 1.2 m before budbreak and maintained well-watered through fruit set. Thereafter, water is applied when the sensors in the top 0.8 m converge on 15 to 17% scaled frequency (with 0% in air and 100% submerged in water), corresponding to a Ψs of −1.0 to −1.2 MPa, while the soil below 0.8 m is permitted to dry down. Water is replenished over the entire soil profile in mid-October to minimize cold injury to the roots. Every time the water is turned on or off, a running total of irrigation water applied to each block is automatically calculated based on emitter spacing, number, and flow rate, as well as volumetric flow rate and pressure transducers.
Berry sampling
A random number generator was used to assign 100 sentinel vines across 17 to 20 rows in each vineyard block, excluding outside rows and the five vines at either end of each row. The sentinel vines were tagged for repeated berry sampling. Samples were collected from the same vines in 2017 and 2018. Two clusters per shoot in both sun-exposed and shaded positions on both sides of the canopy were sampled by manually plucking three to five berries per cluster from the top-front, top-back, middle-front, middle-back, and bottom positions. Three 300-berry replicates were collected in three passes between 0600 and 0900 hr. The first sample from each cultivar consisted of green-hard berries collected during the lag phase (9 Aug 2017 and 2 to 3 Aug 2018), 1 to 2 wk before veraison, which was defined as 50% color change. Samples of 100 berries were then taken for TSS determination once or twice per week until they reached a threshold of ~20 (2017) or 18 Brix (2018); these samples were not used for data analysis as they were not replicated and TSS was the only measured trait. Thereafter, full sampling occurred weekly for 9 (2017) or 10 wk (2018).
Berry analysis
The berry samples were alternately analyzed by a commercial lab (ETS Laboratories) or in-house at the vineyard: 1 wk at the lab (starting with the preveraison berries), 1 wk in-house, and so on. Lab samples were sent inside a cooler by courier for same-day analysis of berry weight, TSS, titratable acidity (TA), pH, total hexoses (glucose + fructose), and malate (L-malic acid). The in-house samples were processed as follows: after weighing, each 300-berry lot was crushed using a rubberized pastry rolling pin while within a plastic bag and strained into a 250-mL beaker; clean juice was pipetted onto a calibrated refractometer (HI96801, Hanna) for TSS determination. Sugar content (i.e., amount per berry) was estimated from berry weight and TSS (Keller et al. 2016). The contents of TA, malate, and hexoses were similarly estimated after conversion of the volume-based concentrations (w/v) to weight-based concentrations (w/w), using juice density (D). The D (in kg/L) was estimated from TSS (in Brix) using a linear approximation for the range 5 to 30 Brix, derived from tabulated data (Tanner and Brunner 1987): D = 0.9944 + 0.0044 × TSS.
Weather data
Daily weather data were obtained from the AgWeatherNet Benton City.E station (https://weather.wsu.edu) that was installed in 1995 at the same elevation and ~50 m from the vineyard block. Growing degree days (GDD) for the period 1 April through 31 Oct were calculated from daily mean temperatures derived from daily maximum (Tmax) and minimum temperatures, using a base temperature of 10°C. The daily mean VPD was calculated from daily mean temperature and relative humidity, as described in Keller (2020).
Data analysis
Data were analyzed using Statistica ver. 14.2 (Cloud Software Group). Effects of cultivar, time (ripening week), and year were analyzed by three-way analysis of variance (ANOVA) and F-test. Interactions were often significant; therefore, ripening data were also analyzed by one-way ANOVA for each year and cultivar. The last Merlot sample in 2017 was removed from the data set since leaf loss and bird damage rendered sampling unreliable. Because the commercial lab data systematically showed 2 to 6% lower berry weights and higher TSS than the in-house data (Supplemental Table 1), the lab data were corrected by the average yearly difference for each cultivar. Rates of berry growth and sugar accumulation were estimated from the changes in berry weight and sugar content between two consecutive sampling dates, using the pooled data from 2017 and 2018. The pH values were converted to H+ concentrations for data analysis, and means were converted back to pH for presentation. Tukey’s honestly significant difference test was used for post-hoc means comparisons when main effects were significant. Associations between key response variables were tested using Pearson correlation analysis. Regression analysis was used to determine the quantitative relationship between hexoses and TSS, and between malate and TA. All other curves were fitted using the distance-weighted least squares procedure.
Results
Weather and phenology
The two growing seasons were very similar in terms of total heat accumulation (1912 GDD in 2017, 1914 GDD in 2018), but warmer than the long-term average at the vineyard site (1755 GDD for 1995 to 2020; Figure 1A). The mean growing season temperature (GST) was 18.8°C in 2017 and 18.9°C in 2018, compared with a long-term average of 18.1°C. While 2017 experienced 90 days with Tmax > 30°C, of which 47 had Tmax > 35°C, there were only 2 days with Tmax > 40°C. In 2018, there were 80 days with Tmax > 30°C, 28 days with Tmax > 35°C, and 6 days with Tmax > 40°C, peaking at 41.4°C during a series of three successive heatwaves between mid-July and mid-August. The average daily mean temperature from fruit set to veraison (dates of phenological stages averaged across cultivars) was 24.9°C in 2017 and 23.4°C in 2018. The average temperature between veraison and harvest was 19.7°C in both years, and even the first 3 wk of ripening differed by only 0.2°C (23.9°C in 2017 and 23.7°C in 2018). The mean daily temperature decreased from nearly 30°C to <10°C over the course of the ripening period, with a pronounced drop after the first week of September in both years (Figure 1B). The decrease in temperature during ripening was also associated with a decrease in daily mean VPD from ~2.5 kPa to <0.5 kPa (Figure 1C).
Growing degree day (GDD, base 10°C) accumulation from April through October (A), and changes in daily mean temperature (T; B) and vapor pressure deficit (VPD; C) during the ripening period in a vineyard in southeastern Washington in 2017 and 2018. Data were obtained from the AgWeatherNet Benton City.E station (https://weather.wsu.edu) located ~50 m from the vineyard.
Annual precipitation varied from 223 mm in 2017 to 150 mm in 2018; the long-term average at this site is 157 mm. Rainfall during the April to October growing season was 72 mm in 2017 and 65 mm in 2018. Virtually no rain fell between fruit set and harvest (15 mm in 2017 and 3 mm in 2018). The total seasonal irrigation water amount was 263 mm for Cabernet Sauvignon and 245 mm for Merlot and Syrah in 2017, compared with 274 mm for Cabernet Sauvignon and 257 mm for Merlot and Syrah in 2018.
The dates of fruit set and veraison (50% color change) differed by 1 to 4 days among cultivars, but Merlot was harvested for commercial use 23 (2017) and 21 (2018) days before Cabernet Sauvignon, with Syrah in the middle (Table 1). Across cultivars, budbreak, fruit set, veraison, and harvest occurred earlier in 2018 than in 2017, but the period from fruit set to veraison lasted 63 days in 2017 and 69 days in 2018 (Table 1). Though the three blocks were harvested for wine production, the 100 data vines were not harvested, to permit continued sampling.
Dates of key phenological stages and harvest (by day of year), harvest yield by block, berry weight, and total soluble solids (TSS) on the sampling date closest to harvest (within 3 days) for three winegrape cultivars in a vineyard in southeastern Washington in 2017 and 2018.
Berry growth and sugar accumulation
Merlot had berries of similar weight in both years, but in 2018, Cabernet Sauvignon berries were 5% heavier and Syrah berries were 14% heavier than in 2017 (Figure 2A). The weight of Cabernet Sauvignon berries increased from 0.6 g before veraison to 1.1 g within 3 wk after veraison in both years, then declined to 0.9 (2017) or 1.0 g (2018) by 10 wk after veraison. Merlot berry weights increased from 0.6 to 1.2 g by 3 wk postveraison before decreasing to 1.1 g by the last sampling date in both years. Syrah berry weights increased from 0.8 to 1.5 g by 3 wk (2017) or to 1.7 g by 2 wk (2018) after veraison, then decreased to 1.3 (2017) or 1.6 g (2018) through week 10 (Figure 2A). Consequently, the three cultivars nearly doubled their berry weight from the late lag phase to the maximum weight 2 to 3 wk postveraison, then gradually lost between 6% and 15% of their weight by the time sampling stopped in mid- to late October. Though a loss of turgor was evident to the touch, no visible symptoms of berry shriveling were detected.
Changes in berry weight (A), total soluble solids (TSS) concentration (B), and sugar content per berry (C) during ripening of three winegrape cultivars in a vineyard in southeastern Washington in 2017 and 2018. Sampling occurred ~1 wk before veraison (50% color change) and then weekly after reaching ~18 Brix. Data show means ± SE if SE exceeds symbol size. Asterisks indicate significant differences (p < 0.05) between years.
TSS concentration was less variable within sampling dates than berry weight and increased more than four-fold during ripening. Berry weight and TSS correlated in all cultivars, showing a broad weight maximum between 21 and 25 Brix (Figure 3A). Hexose concentration (in g/L) was a linear function of TSS (in Brix) with a common regression equation across the three cultivars, taking into account the ~11% increase in juice density as TSS increases from 5 to 30 Brix: [Hexoses] = 12.8 × TSS − 49.4 (Figure 3B). Adding TA to the hexose concentration shifted the intercept closer to zero and decreased the slope of the regression line, which reflects the major contribution of organic acids to TSS prior to veraison. Starting from ~6 Brix in green-hard berries immediately before veraison in all three cultivars, TSS rose rapidly in the first 3 to 4 wk postveraison, after which the rate of increase slowed considerably (Figure 2B). Nevertheless, in all cases TSS continued to increase through the end of the sampling period 10 wk after veraison, substantially exceeding 25 Brix in all cultivars and in both years (Table 2).
Associations between total soluble solids (TSS) and berry weight as a percentage of maximum weight (A); hexoses plus titratable acids (TA) or hexoses alone (B); TA or malate (C); and pH (D) during ripening of three winegrape cultivars in a vineyard in southeastern Washington. Sampling occurred ~1 wk before veraison (50% color change) and then weekly after reaching ~18 Brix. Data for 2017 and 2018 were pooled (all p < 0.001).
Berry weight and composition after extended ripening of three winegrape cultivars in a vineyard in southeastern Washington in 2017 and 2018. DOY, day of year; TSS, total soluble solids; TA, titratable acidity.
Cabernet Sauvignon, the cultivar with the smallest berries in this study, accumulated the least berry sugar, and Syrah, the cultivar with the largest berries, accumulated the most (Figure 2C and Table 2). The amount of sugar per berry (i.e., sugar content) increased most rapidly during the initial ripening phase as both berry size and sugar concentration increased concurrently. Sugar accumulation then began to slow and generally ceased 3 to 4 wk after veraison. On average across the two years, Cabernet Sauvignon, Merlot, and Syrah berries reached 23.5, 22.7, and 23.1 Brix, respectively, by the time the sugar content plateaued. The rates of berry growth and sugar accumulation, estimated for the pooled data from 2017 and 2018, were highest at the beginning of ripening then declined rapidly (Figure 4). Despite some temporal variability, the rate of change in berry weight (i.e., berry growth) of all three cultivars fell below zero ~4 wk after veraison (Figure 4A). The rate of change in sugar content (i.e., sugar accumulation) joined the zero line ~1 wk later (Figure 4B), while the rate of change in sugar concentration approached zero at the same time but remained positive near 0.05 Brix/day (Figure 4C).
Changes in rates of berry growth (A), sugar accumulation per berry (B), and increase in total soluble solids (TSS; C) during ripening of three winegrape cultivars in a vineyard in southeastern Washington. Sampling started ~1 wk before veraison (50% color change), and rates were calculated for the intervals leading up to each week shown on the x-axis. Data for 2017 and 2018 were pooled and show means ± SE if SE exceeds symbol size.
Changes in acidity
Across cultivars and years, pH correlated strongly with the concentration of both TA (Figure 5A) and malate (Figure 5B), and malate correlated with TA (Figure 5C). Both malate and TA also inversely correlated with TSS (Figure 3C). TA concentration was similar across cultivars, but pH throughout the sampling period was lowest in Cabernet Sauvignon and highest in Merlot, though the difference was especially marked before veraison (Figure 6C). The cultivar effect was consistent between years and persisted even when pH was compared at the same TA or malate concentration (Figure 5A and 5B). In all cultivars, pH increased rapidly during the first 2 to 3 wk of ripening, after which the increase slowed temporarily. Rather than reaching a plateau, however, pH increased more readily again during the phase of passive increase in sugar concentration above ~23 Brix (Figure 3D).
Associations between titratable acidity (TA) and pH (A); malate and pH (B); TA and malate (C); and TA and the percentage of malate relative to TA (D) during ripening of three winegrape cultivars in a vineyard in southeastern Washington. Sampling occurred ~1 wk before veraison (50% color change) and then weekly after reaching ~18 Brix. Data for 2017 and 2018 were pooled (all p < 0.001) and followed a similar pattern across cultivars (A to C), or a distinct pattern for Syrah (D).
Changes in malate (A), titratable acidity (TA; B), and pH (C) during ripening of three winegrape cultivars in a vineyard in southeastern Washington in 2017 and 2018. Sampling started ~1 wk before veraison (50% color change) in both years, but subsequent samples (starting at ~18 Brix) were collected 1 wk apart in the two years. Data show means ± SE if SE exceeds symbol size. Asterisks indicate significant differences (p < 0.05) between paired samples across years.
Malate concentration (in g/L) could be well predicted from TA (in g/L) using a common binomial equation across cultivars: [Malate] = 0.01 × [TA]2 + 0.26 × [TA] − 0.2 (Figure 5C). However, Syrah had more malate at any TA compared with Cabernet Sauvignon and Merlot (Figure 5D). In other words, malate contributed more markedly to the TA of Syrah berries at all stages of development (p < 0.001). Malate made up 52, 48, and 58% of the TA of preveraison Cabernet Sauvignon, Merlot, and Syrah berries, respectively. In postveraison berries, malate made up 24, 20, and 37% of the TA of Cabernet Sauvignon, Merlot, and Syrah, respectively. Compared with 2017, all three cultivars in 2018 accumulated less malate before veraison, and the proportion of malate was lower (both before and after veraison) (Figure 6A). Both within and across cultivars, the rate of postveraison malate breakdown correlated strongly with the rate of hexose accumulation (r = −0.90, p < 0.001), indicating that the two rates slowed in concert during ripening. The malate degradation rate was comparable between the two years, decreasing from about −0.35 mg/day to zero, 7 wk after veraison in Cabernet Sauvignon and Merlot (Figure 7). Consistent with its higher sugar accumulation rate, malate degradation proceeded more quickly in Syrah and apparently did not cease until 9 wk after veraison. Yet by the last sampling date, Syrah berries still contained about twice as much malate as Cabernet Sauvignon berries and three times as much as Merlot berries (Table 2). Thus, Syrah berries not only accumulated more sugar but also retained more malate than Cabernet Sauvignon and Merlot berries. Unlike malate, however, TA generally continued to decrease, even if slowly, until the last sampling date (Figure 6B).
Changes in the rate of malate degradation during ripening of three winegrape cultivars in a vineyard in southeastern Washington. Sampling started ~1 wk before veraison (50% color change), and rates were calculated for the intervals leading up to each week shown on the x-axis. Data for 2017 and 2018 were pooled and show means ± SE if SE exceeds symbol size.
Discussion
This study, conducted over two years in Washington’s warm and arid Red Mountain AVA, found that Cabernet Sauvignon, Merlot, and Syrah berries actively accumulated sugar up to a TSS concentration of ~23 Brix, while degrading malate, for 3 to 4 wk after veraison (50% color change). Thereafter, TSS continued to increase passively as the berries began to shrink, losing 6 to 15% of their weight late in the growing season. The rates of berry growth, sugar accumulation, and malate breakdown were most rapid during the initial 2 to 3 wk and declined progressively over the course of ripening. However, active sugar accumulation continued for ~1 wk after the cessation of berry growth, and malate catabolism persisted for at least another week. Compared with Cabernet Sauvignon and Merlot, the bigger berries of Syrah accumulated more sugar (i.e., they reached a higher sugar content despite their lower sugar concentration) and retained more malate (in terms of both content and concentration) by late ripening. Malate also made up a greater fraction of TA in Syrah berries, suggesting their pH was less dependent on other organic acids, namely tartrate, than that of Cabernet Sauvignon and Merlot.
Syrah, like other cultivars with bigger berries (e.g., Suter et al. 2021), accumulated more sugar per berry, likely because bigger berries exert a greater sink strength (Keller et al. 2022). Growth and active sugar accumulation by ripening berries requires import of sucrose and water via the phloem and storage of hexoses in the mesocarp vacuoles (Zhang et al. 2006, Zhang and Keller 2017). Our data suggest that sugar accumulation stopped ~1 wk after berry growth ceased and when TSS had reached 23.5, 22.7, and 23.1 Brix in Cabernet Sauvignon, Merlot, and Syrah, respectively. Lower TSS thresholds of 20 Brix or even 18 Brix were previously published for Cabernet Sauvignon and Syrah (Shahood et al. 2020, Antalick et al. 2021, Previtali et al. 2021), but the threshold for Merlot in the present study is virtually identical to the 22.5 Brix at which the apoplast pH was found to converge with the vacuole pH in Merlot berries (Keller and Shrestha 2014). Subsequent increases in TSS must result from passive concentration via water loss from the berries. Berry dehydration during extended ripening periods is inevitably associated with yield loss. In agreement with our results, an 8-yr field trial with Syrah experienced an average 13% decline in berry weight, and consequently cluster weight and yield, as TSS increased from 24 to 26 Brix (Hunter and Volschenk 2024).
Shahood et al. (2020) asserted that berry growth and sugar accumulation stop simultaneously when a berry reaches its maximum weight. This idea implies one of three possibilities: (i) cessation of phloem import stops berry growth; (ii) growth arrest results in cessation of phloem import; or (iii) another process halts both growth and phloem import. Our results suggest a less strict coupling of the termination of growth and sugar accumulation, allowing for some phloem import to continue for a while after berry growth has stopped. While growth of ripening berries requires water from phloem inflow, the increasing cutinization of, and peroxidase activity in, epidermal cell walls leading to stiffening of the skin may put the upper limit on fruit growth (Thompson et al. 1998, Bargel and Neinhuis 2005), though in our study, this upper limit clearly differed both among cultivars and between years. Phloem import, however, in addition to being dependent on plant water status, should decrease as the temperature cools and the photoperiod shortens, which increasingly limits the canopy’s daily photosynthate production (Tarara et al. 2011). Import might also cease gradually as a berry’s rising sugar concentration in the apoplast approaches that within the vacuoles (Wada et al. 2008, Keller and Shrestha 2014), which would increasingly hinder phloem unloading. Alternatively, gradual or sudden loss of membrane integrity would likewise progressively impede phloem unloading in grape berries even as metabolism, such as malate degradation, persists for a while at a slowing rate (Keller and Shrestha 2014, Keller et al. 2016). Though membrane failure results in cell death, all mesocarp cells do not lose viability at the same time (Krasnow et al. 2008, Hoff et al. 2021). Because the rate and degree of cell viability loss may differ among cultivars and external conditions (Fuentes et al. 2010, Caravia et al. 2015), it might be instructive to test if the amount of residual malate remaining in grape berries after extended ripening correlates with the extent of cell death.
Grape berries will shrink once phloem inflow decreases below the sum of the rates of xylem backflow (which increases as leaf water potential decreases) and berry transpiration (which increases as VPD increases). Berry shrinkage in our study may have been rather slow because the cessation of phloem inflow roughly coincided with the drop in both temperature and VPD that occurred in both years. During active sugar accumulation, berry size is quite insensitive to changes in plant water status (Greenspan et al. 1994, 1996, Keller et al. 2006), and late-season berry shrinkage leading to passive increases in sugar concentration can occur under both well-watered and water-stress conditions (Knipfer et al. 2023), suggesting that VPD may be more important than plant water status in driving berry shrinkage. Shrinkage has been hypothesized to correlate with loss of mesocarp cell membrane integrity (Krasnow et al. 2008, Tilbrook and Tyerman 2008, Fuentes et al. 2010, Caravia et al. 2015), but the cause-effect relationship remains unclear. It is possible that an increasing oxygen shortage during ripening damages cell membranes (Xiao et al. 2018). Alternatively, membrane failure might occur as a consequence of osmotic stress due to the increasing sugar concentration (Bondada et al. 2017). In principle, however, membrane failure is not a prerequisite for berry shrinkage, which may occur simply as a result of a negative water balance, as mentioned above (Zhang and Keller 2015).
Syrah and to a lesser extent, Cabernet Sauvignon, berries grew significantly larger in 2018 than in 2017, though the vineyard received comparable amounts of water in the two years, and yields were also similar. Moreover, bloom time temperatures (average Tmax of 28.7°C in 2017, 29.7°C in 2018) were in the optimal range for fruit set and subsequent berry growth (Keller et al. 2022), and the average temperature during the first 3 wk of ripening was virtually identical in 2017 and 2018. We suspect that temperature differences before and after fruit set may have been important in determining berry growth. Higher early-season temperatures are associated with bigger flowers, which tend to result in bigger berries (Keller et al. 2010). The average Tmax for the budbreak to fruit set period was 24.3°C in 2017 and 26.6°C in 2018. Additionally, the first 4 wk after fruit set comprise the main period for cell division in the berry mesocarp (Keller 2020). The average Tmax for that period was 32.4°C in 2017 and 28.4°C in 2018. Including the following 2 wk, after which cell division in the skin typically ceases as well, the average Tmax was 33.0°C in 2017 and 29.5°C in 2018. The lower Tmax before bloom and the higher Tmax after fruit set in 2017 may have been sufficient to reduce cell division in the flowers and berries relative to 2018 (Kliewer 1977, Keller et al. 2010). Syrah berry growth might be more sensitive to temperature differences than that of Cabernet Sauvignon and Merlot. Such differences are relevant here because both GDD accumulation and GST place the two growing seasons at the high end of a warm climate (Jones et al. 2010).
The negative sugar accumulation rates that we sometimes found during late ripening are probably a sampling artefact. Though small amounts of sugar may be lost from grape berries because of the coupling of apoplastic phloem unloading with xylem backflow, the pedicels effectively retrieve the leached sugar (Zhang et al. 2022). One might argue that this retrieval mechanism could become overwhelmed if sugars were to leak from vacuoles once the mesocarp membranes start failing during the late ripening phase (Krasnow et al. 2008, Caravia et al. 2015). However, no loss of berry sugar was observed in grape berries suffering from a ripening disorder termed “berry shrivel” (also referred to as “sour shrivel”), which is associated with premature membrane failure (Keller et al. 2016).
The preveraison berries of all three cultivars accumulated 23% more malate per berry (34% higher concentration) in 2017 than in 2018, and again the annual variation was greater in Syrah than in Cabernet Sauvignon and Merlot. On average, the fruit set to veraison period was 1.5°C warmer (and 6 days shorter) in 2017 than in 2018. Higher temperatures during early fruit development tend to enhance malate accumulation, while higher temperatures during (early) ripening typically accelerate malate catabolism (Richardson et al. 2004, Sweetman et al. 2014, Hewitt et al. 2023). In our study, the 0.2°C difference between the two years in the average temperature during the first 3 wk of ripening was insufficient to cause differences in malate degradation rates. In both years and in all cultivars, pH continued to increase after sugar accumulation and malate catabolism had ceased; i.e., the free proton concentration decreased even as the berries were shrinking. Given that the TA concentration decreased concurrently with the late-season rise in pH, it is possible that some tartrate degradation might have occurred at that time (Bondada et al. 2017). Though tartrate breakdown is not a universal phenomenon in overripe grape berries (Keller and Shrestha 2014, Hernández-Montes et al. 2021), some oxidation might occur in the presence of iron via the Fenton reaction (Coleman et al. 2023). Changes in mineral nutrients such as potassium seem unlikely to be involved in the late pH increase since import of the phloem-mobile K+ usually ceases concomitantly with sugar import (Rogiers et al. 2006, Liu et al. 2024). It is possible, however, that continued proline accumulation (Stines et al. 1999) might contribute to the rise in juice pH. Proline production from glutamate involves OH− release (Smith and Raven 1979), and proline might serve to protect the cells from hexose-induced osmotic stress.
Conclusion
The present study found that the berries of Cabernet Sauvignon, Merlot, and Syrah winegrapes growing in a warm and arid climate expanded for 2 to 3 wk after veraison and imported, and thus actively accumulated, sugar while breaking down malate for at least another week after berry growth ceased. By the time sugar import stopped, TSS concentration was ~23 Brix. TSS then continued to increase passively as berries began to shrink, losing 6 to 15% of their weight during the late ripening phase. Our results show that sugar concentrations above 23 to 24 Brix can only be achieved by gradual berry dehydration, which is necessarily associated with yield loss and increased risk from diseases and fall frosts. These results have direct practical implications. For example, if a winery wanted to harvest fruit after an extended ripening period and then compensate a grower for the yield loss due to berry shrinkage, they could simply multiply the harvested tonnage with the harvest TSS divided by the threshold TSS at the maximum sugar content. In our study, that threshold was 23.5, 22.7, and 23.1 Brix for Cabernet Sauvignon, Merlot, and Syrah, respectively. Thresholds could be easily adapted to different cultivars and local conditions through repeated and consistent collection of berry samples throughout ripening. As a reference and starting point, we recommend collection of an initial sample of green-hard preveraison berries when the first berries in a vineyard block soften or turn color.
Supplemental Data
The following supplemental materials are available for this article in the Supplemental tab above:
Supplemental Table 1 Average postveraison berry weight and total soluble solids (TSS) measured in-house (CdC) and in a commercial lab (ETS) for berries from three winegrape cultivars grown in a vineyard in southeastern Washington in 2017 and 2018. Data are means ± SE.
Data Availability
The data underlying this study are available on request from the corresponding author.
Footnotes
This work was funded by Ciel du Cheval Vineyard and the U.S. Department of Agriculture’s National Institute of Food and Agriculture (Hatch project 7003737). We thank Jim Holmes for support and stimulating discussions during this project. The authors declare no conflicts of interest in the publication of this work.
Casciato WK and Keller M. 2025. Active and passive changes in sugar concentration during grape ripening. Am J Enol Vitic 76:0760019. DOI: 10.5344/ajev.2025.24069
By downloading and/or receiving this article, you agree to the Disclaimer of Warranties and Liability. If you do not agree to the Disclaimers, do not download and/or accept this article.
- Received December 2024.
- Accepted May 2025.
- Published online July 2025
This is an open access article distributed under the CC BY 4.0 license.
References
Vol 76 Issue 2
Jump to section
Related Articles
Cited By...
More from this TOC section
Similar Articles
