Abstract
Background and goals The water balance of a grape berry determines its final size and has implications for yield and quality. However, how water moves into and out of a berry and how water movement is affected by grapevine water status remains controversial. This synopsis examines the available evidence to develop suggestions for vineyard irrigation management.
Methods and key findings Studies published over the past 20 yr have firmly refuted the notion that ripening grape berries become hydraulically isolated from the vine. Instead, the berries’ high demand for sugar leads to a massive increase in phloem inflow that requires discharge of surplus water via berry transpiration and xylem backflow to the leaves. Although berry size becomes unresponsive to short-term changes in soil moisture, ripening berries will shrink gradually when vines are water-stressed. Conversely, the berries’ propensity to absorb surface water can result in berry splitting, leading to berry expansion or shrinkage, and thus dilution or concentration of berry contents, depending on weather conditions.
Conclusions and significance Unlike rainfall, irrigation during grape ripening does not dilute grape composition–unless it is supplied by overhead sprinklers. This Insight argues that while some water deficit is desirable for high-quality grape production, growers should focus mostly on preveraison water deficit for maximum benefits.
Introduction
Many wine producers, especially makers of red wine, envision an ideal grape cluster to be one with small berries of uniform maturity that are spaced loosely over the rachis framework. While there is disagreement regarding the optimal vineyard management practices that ought to be implemented to produce such clusters, irrigation management is the most powerful strategy that growers can employ in sufficiently dry climates to manipulate berry size and ripening. But although it has long been known that preveraison vine water deficit, not postveraison water deficit, is associated with small berries (Hardie and Considine 1976, Matthews and Anderson 1989, Ojeda et al. 2001, Intrigliolo et al. 2012), many winemakers and regulators still fear that irrigation during the ripening period, and especially close to harvest, will dilute grape quality by increasing berry size. This notion may have its origin in the observation that in climates where summer rainfall is common, high-quality red-wine vintages are often (though not universally) associated with dry years (van Leeuwen et al. 2009).
In the absence of significant summer rainfall, soil drydown will result in unirrigated vines becoming increasingly water-stressed as the growing season progresses. As a consequence, although irrigation is becoming more commonplace in Europe where sufficient water is available, regional rules and regulations often restrict, or outright forbid, water application during the grape ripening phase. Moreover, many field trials, especially those conducted in dry European regions, have applied various deficit irrigation regimes that remained unchanged (e.g., as constant fractions of vineyard evapotranspiration) from treatment start through harvest (e.g., Romero et al. 2022). Like lack of summer rainfall, this approach also leads to progressively intensifying vine water stress, because replenishing <100% of the water that evaporates from a vineyard necessarily leads to soil drying. While it has been undisputed for at least the last 2000 yr that some degree of water deficit is beneficial for vigor control and grape and wine quality (Columella, c. 70 ad), the ideal severity and timing of the stress remain areas of research and debate. Unfortunately, however, even recent reviews and meta-analyses have considered water deficit severity only as a seasonal average instead of by phenological stage (Gambetta et al. 2020, Cameron et al. 2024, Uriarte et al. 2025).
This short paper reviews the scientific evidence underlying water movement in grapevines, with special emphasis on grape berries, exposing some myths in the process. It explains how water moves into and out of berries and how some of this water contributes to the determination of final berry size and composition. The mini-review then looks at key differences between the effects of certain irrigation practices and rainfall, and their implications for berry size and composition. Although it is not intended to be a comprehensive review of irrigation approaches and effects, the paper concludes with some general recommendations for vineyard irrigation management.
Water flow: From roots to leaves and beyond
Water flow from the soil to the leaves of grapevines, as in other plants, generally follows what we call the soil-plant-atmosphere continuum (Tyree and Zimmermann 2002). In this widely accepted model, soil water is taken up at the root surface and flows along the apoplast (cell walls and intercellular spaces) or symplast (from cell to cell across interconnecting pores termed plasmodesmata) to the endodermis. There, the waterproof Casparian strip forces the water through the symplast before it is released into the water-conducting vessels of the xylem for transport to the canopy, following a gradient of negative hydrostatic pressure generated by the evaporation of water from the leaves to the atmosphere. The rate of water flow (F) in general can be described in terms of the ratio between a driving force generating the flow and a resistance (or series of resistances) opposing it. Simply put, the driving force is the difference or gradient in water potential (ΔΨ) between two points, and the resistance consists of the hydraulic constraints (rh) imposed by the length and layout of the pathway between these points; thus, F = ΔΨ/rh.
While most water flowing up a vine’s trunk evaporates in the process of transpiration, some is retained for cell expansion by the growing tissues, and some moves to the phloem to enable transport of sugar and other nutrients to those tissues and to storage organs (Keller 2025). In the leaves, transpiration occurs predominantly across specialized and tightly controlled valves termed stomata. Though grape flowers also possess stomata, these are mostly lost during capfall at bloom; those that remain gradually become dysfunctional as the berries develop after fruit set (Blanke and Leyhe 1988, Rogiers et al. 2004). Consequently, the transpiration rate per unit surface area of berries (0.01 to 0.1 mmol H2O/m2sec) is one to two orders of magnitude slower than that of leaves (1 to 10 mmol H2O/m2sec) (Zhang and Keller 2015). As leaves are designed to maximize CO2 uptake for photosynthesis, evaporative water loss is an unavoidable side effect. Grape berries, however, are designed to minimize water loss or gas exchange, so much so that their interior often suffers from oxygen deficiency (Xiao et al. 2018).
Berry shrinkage and the power of sinks
Unlike leaves, grape berries have no stomatal control to regulate their transpiration rate, which is not only slow but also tracks the surrounding air vapor pressure deficit (VPD) rather precisely (Zhang and Keller 2015; Figure 1). Hot and dry air thus drives evaporative water loss from berries, while cool and moist air hinders such water loss. Their poor capacity for evaporative cooling also makes berries much more vulnerable to heat injury, and hence sunburn, than leaves (Gambetta et al. 2021, Müller et al. 2023). Nevertheless, water loss will lead to berry shrinkage on water-stressed grapevines (Greenspan et al. 1996, Keller et al. 2006). Though shrinkage can occur both before and after veraison, preveraison berry shrinkage is much more severe at the same level of water stress. But unlike postveraison shrinkage, preveraison shrinkage may be reversible upon replenishment of soil moisture (Keller et al. 2006). Additionally, as they begin ripening, shrinking berries start to re-expand, accumulate sugar, and change color even in the face of severe water stress leading to leaf shedding (Keller et al. 2015; Figure 2A). The berries are such strong sinks at this time that they will attract sugar from the vine’s perennial storage pools when none is available from the leaves (Candolfi-Vasconcelos et al. 1994). Replenishing soil moisture during this initial ripening phase accelerates both berry expansion and sugar accumulation as photosynthesis of any remaining leaves recovers rapidly (Keller et al. 2015). In susceptible cultivars, such recovery irrigation following severe drought stress may even result in berry splitting (Chang et al. 2020). Girdling of the peduncle (i.e., phloem removal; Figure 2B), but not surgical removal of the peduncle xylem (Figure 2C), prevents berry expansion following rewatering. Clearly, phloem inflow increases berry size and sugar content in both water-stressed and non-stressed vines at veraison, while the contribution of xylem inflow is negligible.
The transpiration rate of Syrah grape berries tracks the atmospheric vapor pressure deficit (VPD). Figure modified from Zhang and Keller (2015).
Preveraison Concord grape berries shrink under plant water stress and re-expand at the onset of ripening despite continued stress (A). Peduncle girdling (B) but not xylem removal (C) prevents re-expansion of shrunk berries upon rewatering. Figure modified from Keller et al. (2015).
Xylem flow reversal, not hydraulic isolation
It used to be widely accepted that the decline in xylem inflow at the onset of ripening resulted from renewed berry expansion rupturing the berry’s xylem vessels, leading to its hydraulic isolation from the rest of the plant (Düring et al. 1987, Findlay et al. 1987). Beginning in 2005, evidence began to accumulate demonstrating that the xylem remains intact and functional in ripening berries (Bondada et al. 2005, Chatelet et al. 2008, Choat et al. 2009, Tilbrook and Tyerman 2009, Knipfer et al. 2015). It turns out that water flow in the berry xylem undergoes a gradual and sustained reversal of flow direction at ripening onset (Keller et al. 2006, Zhang and Keller 2017). Sugar import via the phloem brings with it a surplus of phloem water and, as the stiffening skin increasingly restricts berry expansion, only a small portion of this water is used for berry growth. The remainder is discharged via berry transpiration and xylem backflow to the leaves (Keller et al. 2015, Zhang and Keller 2017).
Xylem backflow as a strategy for surplus water disposal potentially opens the door for sugar to be swept away from a berry as it switches from symplastic phloem unloading to apoplastic unloading at ripening onset to enable sugar accumulation (Zhang et al. 2006). During apoplastic unloading, sucrose is exported from the phloem to the berry cell walls and is mostly broken down into glucose and fructose. Though these hexoses are readily imported into the berry cells for storage in the vacuoles, some accumulate in the cell walls during ripening (Wada et al. 2008, 2009, Keller and Shrestha 2014). As phloem water follows the unloaded sugar osmotically, it may leach some of the apoplastic sugar back to the xylem. Yet, tracing experiments with 13C-labeled glucose showed that sugar loss is small and any leached sugar is retrieved from the xylem in the pedicel (Zhang et al. 2022). The salvaged sugar may then be recycled back to the berry via the phloem or stored in the pedicel. As sugar accumulation typically triggers anthocyanin biosynthesis in grapevine tissues (Keller 2025), this mechanism also offers a plausible explanation for why grape pedicels sometimes turn red.
Another corollary of the above findings is that they challenge the basic assumption underlying the method introduced by Lang and Thorpe (1989) for estimating xylem and phloem flows in grape berries, and which has since been applied to diverse types of fleshy fruit. While girdling of the pedicel does eliminate phloem inflow into a berry, it also eliminates the main water source and driving force for xylem backflow, making any estimates of vascular flows based on this technique unreliable.
Rapid berry transpiration and xylem backflow help explain why sugar accumulates quickly under warm and dry conditions (high VPD), provided that vines are not overly water-stressed (Zhang and Keller 2017, Pascual et al. 2022). Rates of up to 1 Brix/day during the first 2 wk of ripening were found in warm and arid southeastern Washington, whereas maximum rates of ~0.5 Brix/day were estimated for the same period in cooler and wetter upstate New York (Keller et al. 1998, Casciato and Keller 2025). Thus, xylem backflow also may serve as a backup mechanism to enable continued, albeit slower, sugar accumulation under humid conditions (low VPD) that hinder transpiration. Indeed, blocking either water disposal pathway slows berry sugar accumulation (Zhang and Keller 2017). Moderate and severe water stress, however, curtails canopy photosynthesis and sugar export via the phloem to the berries (Tarara et al. 2011).
Water transport, berry size, and dilution
The transition at ripening onset from the xylem to the phloem as the predominant water import pathway for grape berries means that the berries become much less responsive to changes in soil moisture (Greenspan et al. 1996, Keller et al. 2006, Scholasch and Rienth 2019, Gambetta et al. 2020). Though berries on water-stressed vines will still shrink during ripening, they do so much more slowly compared with preveraison berries. Conversely, although applying a positive pressure of 1 MPa to the root system of fully irrigated pot-grown grapevines leads to copious guttation (water efflux) from the leaf margins, only preveraison berries will expand with such extreme overwatering (Keller et al. 2015; Figure 3). Once the berries have begun accumulating sugar (total soluble solids [TSS] > 10 Brix), their stiffening skin resists expansion under root pressurization while still accommodating berry growth for a while. However, because the skin’s fracture resistance drops rapidly as the berries begin to soften, ripening berries of susceptible cultivars may split under pressure (Zhang and Keller 2017, Chang et al. 2019).
Preveraison, but not postveraison, Merlot grape berries expand when the root system of fully irrigated vines is pressurized at 1 MPa. Inset shows guttation from the leaf margins. Figure modified from Keller et al. (2015).
In a vineyard, berries can absorb surface water during or after rainfall or overhead sprinkler irrigation. In a greenhouse or in a table grape vineyard with plastic cover and undervine microsprinkler irrigation, high nighttime humidity may prevent berry transpiration even if the berries do not absorb water vapor. In both cases, the ensuing increase in the internal pressure on the skin can initiate splitting from existing microcracks (e.g., on the suberized scars where the flower cap used to be attached), which act as stress concentrators (Chang and Keller 2021). Once a berry has split and the flesh is exposed, it becomes highly susceptible to pathogen infection, and it will shrink in dry weather or expand in humid weather. Additionally, continued presence or renewal of surface water may leach out berry solutes and rapidly reduce TSS. Therefore, the dilution of berry composition that growers often observe after rainfall likely results from berry splitting and sugar leaching, rather than from expansion of intact berries.
From theory to practice, and from vineyard to winery
It may be objected that the discoveries and principles discussed in the preceding sections might sound reasonable in theory but may not hold up in real-world vineyard scenarios. This objection has been put to the test independently in dry regions around the world. Summarizing many years of field trials in Australia, Dry et al. (2001) concluded that “to control vegetative and reproductive growth it is recommended that water stress be limited to the period after fruit set”, and “the inappropriate use of water stress, for example during ripening in the belief that maturity will be enhanced, should be avoided.” In both wine and (more heavily cropped) juice grapes in various dry regions, berry size at harvest was related to differences in irrigation water supply before, but not after, veraison (Matthews and Anderson 1989, Intrigliolo et al. 2012, Levin et al. 2020, Keller et al. 2023). While more water before veraison increases berry size, more water during ripening may only alleviate (but not reverse) berry shrinkage. Consequently, errors in irrigation management before veraison cannot be corrected by making changes in postveraison irrigation. Because evapotranspiration from vineyards generally peaks during the preveraison period when the canopy is fully grown and days are long and warm, this is also the period during which deficit irrigation can achieve the greatest water savings (Keller et al. 2016, 2023).
We may take it for granted that overwatering should be avoided in vineyards at any time. Yet grape and wine quality likewise seems to benefit more from preveraison water deficit than from postveraison deficit, though there are exceptions (e.g., Malbec [Shellie 2019] or Gewürztraminer [Kovalenko et al. 2021]). Working with container-grown vines, Palai et al. (2022b, 2023) found that preveraison water deficit led to higher concentrations of aroma volatiles in Merlot and Sangiovese berries than did postveraison water deficit. In a field trial with Viognier, limiting moderate water stress to the preveraison period tended to increase free monoterpenes in the berries (Wang et al. 2019). Even differences in preveraison vine water status as small as 0.2 MPa midday leaf water potential led to distinct aroma volatile profiles of Riesling wine (Diverres et al. 2024). Additionally, in field trials with Cabernet franc, Cabernet Sauvignon, and Tempranillo, preveraison water deficit increased berry or wine anthocyanins more than did postveraison water deficit (Matthews and Anderson 1988, Castellarin et al. 2007, Intrigliolo and Castel 2010, 2012). Reducing irrigation only during ripening of Cabernet Sauvignon led to the least desirable grape and wine composition in terms of anthocyanins and tannins compared with other irrigation regimes, and the same effect occurred with moderate postveraison water stress in Pinot noir (Casassa et al. 2015, Kar et al. 2025). Finally, wine color and the concentration of tannins, polymeric pigments, and other phenolics increased as berry size decreased in Cabernet Sauvignon and Zinfandel (Gil et al. 2015, Casassa et al. 2025).
Providing generally applicable recommendations for vineyard irrigation management is challenging (e.g., Mirás-Avalos and Araujo 2021, Romero et al. 2022). Nevertheless, integrating the findings discussed here, it is recommended that growers focus mostly on the preveraison period when implementing deficit irrigation. Prebloom water stress should be avoided to sustain crop yield in the current and following years. Depending on the intended use of the grapes being grown, the deficit imposed after fruit set may vary from mild for white-wine, table, raisin, and juice grapes to moderate or severe for red winegrapes (for definitions, see Scholasch and Rienth 2019). Some caution may, however, be prudent regarding water stress during the lag phase of berry development; while it may increase aroma volatiles, suddenly alleviating moderate to severe stress at veraison has been associated with the berry shrivel ripening disorder and reduced anthocyanin accumulation (Palai et al. 2022a, 2023, Wenter et al. 2025). Continuing mild water stress through harvest prevents untimely shoot growth and facilitates berry ripening, but more severe stress is not advisable as dehydration does not enhance berry quality. Conversely, when surface water absorption associated with rainfall leads to berry splitting, delaying harvest may mitigate its adverse effects so long as the berries remain healthy.
Conclusion
A grape berry’s final size depends on the balance between water inflow and water outflow during berry development. Water inflow occurs mostly via the xylem before veraison but almost exclusively via the phloem during ripening. Indeed, discharge of surplus phloem water via berry transpiration and xylem backflow is key to rapid sugar accumulation, which is a major reason why berries ripen rapidly in warm and dry environments. Though the size of ripening berries (TSS > 10 Brix) becomes unresponsive to short-term fluctuations in soil moisture, the berries will shrink gradually when vines are water-stressed. Conversely, the drop in the skin’s fracture resistance during berry softening coupled with absorption of surface water can lead to berry splitting. Split berries may expand (if rain continues), shrink (if rain is followed by dry weather), or become infected by pathogens, all of which change grape composition. Unlike surface or drip irrigation, therefore, rainfall or overhead irrigation may dilute or concentrate berry contents, depending on weather conditions. But to manipulate berry size and composition, growers should focus on the preveraison period for maximum benefit. As ripening starts, growers can then decide whether to alleviate the water deficit fully or partially, depending on prevailing weather conditions. However, compensating for inadequate preveraison water management is not possible during ripening.
Data Availability
The data underlying this study are available on request from the corresponding author.
Footnotes
This work was funded by the USDA National Institute of Food and Agriculture (Hatch project 1000186), the USDA Northwest Center for Small Fruits Research (grant number 59-5358-8-284), and the Washington State Grape and Wine Research Program. Conference travel was supported by the American Society for Enology and Viticulture. I thank Lynn Mills and Alan Kawakami for skilled technical support, and the graduate students and postdoctoral scientists whose work is cited in this article.
Keller M. 2026. Vineyard irrigation: A balancing act with consequences. Am J Enol Vitic 77:0770005. DOI: 10.5344/ajev.2025.25035
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 July 2025.
- Accepted November 2025.
- Published online February 2026
This is an open access article distributed under the CC BY 4.0 license.
References
Vol 77 Issue 1
Jump to section
Related Articles
Cited By...
More from this TOC section
Similar Articles
