Physiological Responses Consistent with Near-Isohydric and Anisohydric Behaviour in Grapevine Cultivars Șarba and Fetească Neagră (Vitis vinifera L.) Under Semi-Arid Conditions

Highlights

What are the main findings?

•

We identified a near-isohydric, water-conserving strategy in Șarba, characterised by early stomatal closure at véraison and significantly higher water use efficiency compared to Fetească neagră.

•

We identified an anisohydric, gas exchange sustaining strategy in Fetească neagră, which maintained higher net photosynthesis throughout the season but was associated with premature basal leaf senescence—to the best of our knowledge, the first formal field documentation of this phenomenon for the cultivar.

•

We resolved the two cultivars into two distinct physiological groups, consistent with water-conserving/near-isohydric and gas exchange sustaining/anisohydric behaviour, using PCA, Heatmap Analysis, and linear mixed-effects modelling, with divergence most pronounced at véraison.

What is the implication of the main finding?

•

These findings support cultivar-specific irrigation management: regulated deficit irrigation for Șarba to sustain carbon assimilation, and combined elicitor-based treatments with supplemental irrigation for Fetească neagră.

•

They provide a physiological framework for evidence-based varietal selection and drought adaptation strategies under the increasingly semi-arid conditions projected for Romanian and broader European viticultural regions.

Abstract

Climate change strongly impacts the physiological processes in the grapevine. Two Romanian autochthonous cultivars, Șarba and Fetească neagră (Vitis vinifera L.), grown under semi-arid conditions in 2024 (De Martonne Index: 23.6; Hydrothermal Coefficient: 0.75), showed different physiological behaviours. Leaf gas exchange parameters (A, g_s, E, C_i) and water use efficiency (WUE) were assessed at three phenophases (flowering, berry growth, and véraison), while photosynthetic pigment indices (Chl a/b, Chl/C+X) and leaf dry matter content (dm) were additionally determined at harvest maturity. Multivariate analysis consistently separated the two cultivars into distinct physiological groups: responses consistent with near-isohydric behaviour in Șarba and anisohydric behaviour in Fetească neagră, based on gas-exchange parameters interpreted within the established isohydric/anisohydric framework, as direct water potential measurements were not performed. Șarba exhibited a water-conserving strategy at véraison—characterised by early stomatal closure, high WUE, and maintained chlorophyll—protecting vine water status at the cost of reduced leaf carbon assimilation. Fetească neagră, by contrast, kept its stomata progressively open, sustaining high gas exchange rates in a pattern consistent with progressively declining shoot water potential. Berry sugars are expected to concentrate passively late in ripening, while severe deficit risks berry shrivelling and, under prolonged drought, premature senescence of basal leaves, to our knowledge, a phenomenon not previously reported for this cultivar. These findings support cultivar-specific management, with implications for irrigation scheduling and varietal selection under climate change.

1. Introduction

Grapevine cultivation occupies remarkably diverse climatic zones, that strongly influence the plant physiology and phenological development; climate is recognised as a more decisive factor governing grape composition than soil or variety [1,2]. The climate disruptions of recent decades have left few agricultural sectors untouched [3]. Rising temperatures, erratic precipitation, and the growing frequency of heatwaves and droughts [4] have already compromised berry development and quality [5,6,7], resulting in wines that lack typicality, carry reduced titratable acidity, and accumulate excessive sugars that limit ageing potential [8,9,10,11]. Recent experimental evidence indicates that temperature increases advance all phenophases by about two weeks and reduce stomatal conductance, thus reducing both yields and the compositional balance of the fruit [12]. In response, viticulture research has shifted toward mitigation, above all, varietal selection [13,14,15,16,17,18]: replacing traditional cultivars with heat- and drought-tolerant alternatives [19] or introducing late-ripening varieties that push harvest back by up to two weeks [20,21].

Water scarcity is managed by Vitis vinifera L. via a series of interactive mechanisms including minimisation of transpiration, exploitation of deep-water reserves and cellular osmotic adjustment to manage deficit [22]. Drought remains the principal abiotic constraint on grapevine productivity and berry quality across Mediterranean and continental climates, interfering with every level of vine physiology from stomatal gas exchange to berry development. The response of grapevines to progressive soil water deficit has been characterised in detail over the last two decades, evidencing a complex and genotype-dependent interplay of hydraulic, hormonal and metabolic mechanisms [7,23].

One of the major concepts in grapevine drought physiology concerns how cultivars regulate stomata as soil and plant water potential decline—a classification that, though often framed as a simple binary, is more accurately understood as a continuum: from water-conserving, near-isohydric behaviour at one extreme to gas exchange sustaining, anisohydric behaviour at the other. Isohydric cultivars, like Grenache or Montepulciano prioritise water status over carbon assimilation, closing stomata early to hold leaf water potential roughly constant. Anisohydric cultivars such as Syrah, Sangiovese or Chardonnay take the opposite approach, keeping stomata open to sustain photosynthesis and transpiration even as shoot water potential falls progressively negative [24,25]. Since the same cultivar can change its behaviour toward isohydric under high vapour pressure deficit or depending on the rootstock used, this classification has been further reframed as a plant–environment interaction rather than a fixed hydraulic trait [25,26]. Although widely used, this classification has been refined in recent years. Several studies have shown that a cultivar’s behaviour is not fixed but rather depends on the rate and severity of stress imposition, soil water-holding capacity, vapour pressure deficit, and season, placing varieties along a continuum rather than in distinct categories [26,27,28]. Empirical re-examination of canonical benchmark varieties confirmed this complexity: studies testing Grenache, Syrah, Malbec and Chardonnay under standardised conditions found that the widely accepted classifications did not hold consistently across climatic and edaphic settings, with Malbec occupying a near-isohydric position and Syrah behaving near-isohydrically in some experimental conditions [29]. Even the adaptive value of each response is contested: isohydric behaviour protects hydraulic integrity at the risk of carbon starvation, while anisohydric behaviour sustains carbon assimilation with the cost of hydraulic vulnerability [7,30]. Moreover, recent transcriptomic studies revealed distinct molecular determinants of iso- and anisohydric responses: isohydric genotypes displayed a more rapid upregulation of ABA-related and heat shock protein genes upon imposition of water deficit, while anisohydric genotypes preferentially induced genes for reactive oxygen species-scavenging enzymes and abiotic stress [31,32].

The leaf gas exchange parameters net photosynthesis (A), stomatal conductance (g_s), transpiration (E) and intercellular CO₂ concentration (C_i) are the main indicators of the vine’s instantaneous water use strategy (WUE) and have been widely used to characterise cultivar-specific drought responses [33,34]. Water use efficiency (WUE), the ratio of carbon assimilation to water loss, combines these parameters and indicates the efficiency of resource use under limiting conditions [26]. On the biochemical side, the ratio of total chlorophylls to carotenoids and xanthophylls (Chl/(C+X)) tracks the balance between light harvesting and photoprotection under field drought [35]. Field-grown grapevines generally withstand photoinhibition even under severe stomatal closure [36]; the shifts observed in pigment ratios reflect photoprotective acclimation rather than photosystem damage [35]. Leaf dry matter content (dm) rounds out the picture as a structural trait: inversely correlated with specific leaf area, it complements dynamic stomatal regulation and signals constitutive drought avoidance through denser tissues and reduced cuticular conductance [33].

Drought stress does not act uniformly across the growing season, and its physiological consequences depend heavily on when it strikes. During berry growth, cell division and expansion proceed rapidly, and water deficit at this stage irreversibly reduces berry cell volume and pericarp wall extensibility; subsequent rehydration cannot undo the damage [7,23].

At véraison (the onset of berry ripening), the berry’s principal water supply shifts from xylem to phloem, which by this stage accounts for over 80% of water influx; the xylem, though structurally intact, is no longer the dominant route [37,38]. Surplus phloem-derived water can be recycled back to the shoot through xylem backflow [38,39,40], while phloem unloading shifts to an apoplasmic pathway that enables rapid mesocarp sugar import [38,41].

Véraison is also a window of heightened vulnerability: abrupt changes in vine water status—whether from sudden deficit or rapid rehydration—can trigger berry shrivel disorder, an irreversible arrest of sugar accumulation linked to premature mesocarp cell death and vascular dysfunction [42,43]. During ripening, the berry’s water balance reflects the net difference between phloem and xylem influx and efflux through transpiration and xylem backflow. Under severe deficit, the increasingly negative shoot water potential amplifies backflow, while reduced phloem supply drives progressive berry dehydration, with varietal differences in pedicel and brush-region hydraulic conductance determining cultivar vulnerability [37,38,44]. Moderate water deficit, however, can lead to desirable passive concentration of sugars, anthocyanins and phenolic compounds in the berry, which is the principle underpinning regulated deficit irrigation (RDI) strategies in premium wine production [23,45].

Gambetta et al. [7,46] described grapevines’ complex drought responses, focusing on trait interactions that may be useful in identifying adapted varieties, keeping in mind that these can vary depending on the severity and timing of the water deficit. Above ground, the vine progresses from maximising transpiration under mild stress, through stomatal regulation under moderate stress, to embolism resistance and water loss minimisation under severe deficit. Below ground, the shift runs from exploiting available soil water, through maintaining hydraulic conductance, to ultimately disconnecting from the soil under extreme conditions. Taken together, these traits trace the vine’s trajectory from declining productivity toward mortality risk.

Recent studies indicate that a significant warming trend has affected the majority of Romania’s viticultural regions, profoundly influencing grapevine phenology, physiological responses, and vegetative growth, as well as grape productivity and quality [47,48,49,50,51,52,53,54]. The country’s major vineyards also recorded an average air temperature increase of 0.7 to 2.1 °C, with the Dealu Mare, Târgu Bujor, and Murfatlar vineyards showing the largest variations, indicating a trend toward aridification [55,56].

The aromatic profile and berry composition of autochthonous cultivars, including Șarba, are particularly sensitive to these thermal and hydric shifts during ripening, with documented changes in wine typicity and quality across Romania’s major appellations [57]. The socio-economic impact of climate change may be even greater for wine grapes than for other crops, since the profitability of viticulture is primarily dependent on the relationship between wine quality and the pedoclimatic conditions of the vine-growing sites [58,59]. While grapevines can deploy adaptive responses to hot, dry summers, survival increasingly comes at the expense of yield and quality [7,23]. Water deficit reduces the vine canopy’s efficiency, especially in terms of gas exchange and net assimilation, underscoring the importance of mitigation techniques such as supplemental irrigation [45,60].

In this regard, research on autochthonous grapevine varieties is especially promising. Drought-adaptive traits that international benchmark varieties lack may be present in native cultivars, which have been shaped by centuries of natural selection under local pedoclimatic conditions. Understanding the structural and functional specificities of individual cultivars is required for the successful identification and selection of cultivars able to cope with the conditions imposed by climate change [61].

According to recent research, local cultivars from drought-prone areas show significant variability in hydraulic trait combinations, WUE, and stomatal regulation strategies. Some of these cultivars combine high photosynthetic efficiency with conservative water use in ways that are not fully captured by a strict isohydric or anisohydric classification [26,27], reinforcing the view that these strategies occupy positions along a continuum rather than fixed categories. However, there is still a lack of systematic ecophysiological characterisation of drought response across phenophases for Romanian autochthonous cultivars. This gap is becoming more difficult to justify in light of the documented trend of aridification in Romanian viticultural regions. Existing iso/anisohydric physiology research has focused predominantly on benchmark international cultivars such as Grenache, Syrah, Chardonnay, Sangiovese and Montepulciano; comparable ecophysiological characterisation of Romanian autochthonous cultivars under field drought conditions is, to our knowledge, still lacking.

To address this gap, the current study compares and characterises the physiological drought responses of two Romanian autochthonous cultivars, Fetească neagră and Șarba, across phenophases from flowering through harvest maturity, under semi-arid conditions. Using a multivariate approach integrating leaf gas exchange parameters, water use efficiency (WUE), and assimilatory pigment parameters, we aimed to:

(i)

Identify the water use strategy of Șarba and of Fetească neagră, and its dependence on phenophase;

(ii)

Characterise the leaf structural correlates (stomatal density and leaf dry matter content) and physiological components of drought tolerance in each variety;

(iii)

Discuss the implications of the observed results for berry water balance, shrivel risk, and vineyard management of these varieties under increasing summer drought.

2. Materials and Methods

2.1. The Experimental Site and Biological Material

The trial was conducted in 2024 in the ampelographic collection of the Faculty of Horticulture, University of Agronomic Sciences and Veterinary Medicine Bucharest, Romania (Lat. N: 44°47′07″; Long. E: 26°07′28″; alt. 87 m), on the varieties Șarba (VIVC 10738) and Fetească neagră (VIVC 4120) (Vitis vinifera L.) (Figure 1).

The vineyard is located on a plane surface with reddish preluvosol soil, and the vines were spaced at 2.2 m between rows and 1.2 m (density of 3787 plant ha⁻¹) within the North–South oriented rows. The vines are trained as Guyot on half stem with a mixed pruning system and loading of 28 buds per vine. The vines were over 20 years old at the time of the study. No irrigation was performed in the vineyard during the vegetation season when the determinations were made. Phytosanitary treatments to control diseases and pests were applied in accordance with local standard practice. Soil management provides for natural ground cover.

Șarba is a semi-aromatic white wine variety homologated in 1972, obtained at the Odobești Viticulture and Winemaking Research Station from Italian Riesling by open fertilisation [62]. Molecular analyses have confirmed Italian Riesling and Muscat de Hamburg as parental genotypes [63]. However, Muscat de Hamburg was not historically cultivated in this area, whereas Tămâioasă Românească was a common local variety. Wine volatile profiles and historical distribution records therefore suggest Tămâioasă Românească as the more plausible pollen parent [21,64]. The variety is cultivated in the warmer vineyards of Moldova (Bujorului Hills, Odobești, Panciu, Cotești, Zeletin) and Muntenia and Oltenia (Dealu Mare, Ștefănești, Drăgășani) [65].

Fetească neagră is the most renowned Romanian autochthonous red wine variety, originating from a popular selection of Vitis vinifera L. ssp. sativa (DC.) Hegi, with its place of origin in the Prut River valley near Uricani, Iași County [65,66]. It is cultivated across all major Romanian wine regions.

2.2. Climate Data Measurements

Climatic data were obtained from the Bucharest–Băneasa meteorological station (44°43′ N; 26°10′ E). The 30-year reference baseline (1991–2020) was compared with 2024 observations. Parameters included mean, minimum, and maximum temperatures, and total precipitation. Bioclimatic indices calculated: Hydrothermal Coefficient (HC) [67], Huglin Heliothermal Index (HI) [68], Winkler Index (WI) [69], De Martonne Aridity Index (DMI) [70], and Cool Night Index (CNI) [71]. For 2024, useful heat balance (UHB) was additionally calculated for each phenophase interval, as the sum of degree-days above 10 °C (UHB = Σ(Tmean,i − 10) for days where Tmean,i > 10 °C), accumulated from the end of the preceding phenophase to the date of the phenophase concerned.

2.3. Parameters Determined

2.3.1. Phenological Data

Phenophases (budburst, flowering, véraison, harvest maturity) were recorded following the OIV Descriptor List, 2nd edition [72], using the BBCH scale. Stages were recorded at 50% occurrence (BBCH 008—budburst; BBCH 605—flowering; BBCH 801—véraison; BBCH 809—harvest maturity).

2.3.2. In Situ Leaf Gas Exchange Parameters

Net CO₂ assimilation (A, μmol CO₂ m⁻² s⁻¹), transpiration rate (E, mmol H₂O m⁻² s⁻¹), stomatal conductance (g_s, mol H₂O m⁻² s⁻¹), and intercellular CO₂ concentration (C_i, μmol CO₂ mol⁻¹) were measured in situ on fully expanded, sun-exposed leaves at nodes 4–7 of the main shoot, between 09:00 and 11:00 h under ambient conditions, using a portable infrared gas analyser (LCPro-SD, ADC BioScientific Ltd., Hoddesdon, UK). Measurements were performed at flowering, berry growth (pea size), and véraison. Water use efficiency (WUE) was calculated as the instantaneous ratio A/E. The same tagged leaves at nodes 4–7 were monitored longitudinally across all three phenophases (flowering, berry growth, and véraison). Ambient conditions during measurement campaigns were maintained near natural field levels: photosynthetically active radiation (PAR) approximately 900–1400 µmol m⁻² s⁻¹, leaf chamber temperature approximately 26–32 °C, and ambient CO₂ concentration approximately 410–430 ppm (consistent with the Bucharest regional baseline of ~420 ppm); vapour pressure deficit was not recorded.

2.3.3. Photosynthetic Pigments and Leaf Dry Matter

Chlorophyll and carotenoid content was quantified spectrophotometrically at 663, 646, and 470 nm (Helios Alpha 9423 UVA 1002E, Thermo Scientific, Cambridge, UK) using 80% acetone (diluted from acetone, ≥95.5%, p.a., ACS, ISO grade, Sigma-Aldrich, St. Louis, MO, USA) as solvent, following Lichtenthaler [73]. Results were expressed as mg 100 g⁻¹ FW. Determinations were performed at flowering, berry growth, véraison, and harvest maturity. Leaf dry matter content (dm) was determined gravimetrically by weighing fresh leaf discs before and after oven-drying (Heraeus Instruments GmbH, Hanau, Germany) at 70 °C for 48 h, expressed as (dry mass/fresh mass) × 100.

2.3.4. Stomatal Analysis Method

Stomatal density was determined by the epidermal impression method [74]. A thin uniform layer of colourless nail varnish was applied to the abaxial leaf surface between main veins and allowed to dry. The film was lifted using transparent adhesive tape and mounted on a glass slide with the imprint surface upward. Stomata were counted under a light microscope (Leica DM1000 LED, Leica Microsystems, Wetzlar, Germany) at ×400 magnification in five randomly selected fields of view per leaf. Stomatal density (mm⁻²) was calculated from the mean count and the calibrated field of view area. A minimum of three leaves per cultivar were analysed.

2.3.5. Harvest Quality Parameters

At harvest maturity, the following parameters were determined: grape cluster weight (g), weight of 100 berries (g), soluble solids (°Brix, Atago digital refractometer, Atago Co., Ltd., Tokyo, Japan), titratable acidity (g L⁻¹ tartaric acid, potentiometric titration with 0.1 N NaOH to pH 7.0, TitroLine EASY, Schott Instruments, Mainz, Germany), pH, and polyphenolic potential. For titratable acidity, 0.1 N NaOH was prepared from pelletized NaOH (Labkem, Labbox Group, Mataró, Spain) and standardised with potassium hydrogen phthalate (p.a., Labkem, Labbox Group, Mataró, Spain). The polyphenolic potential was assessed by the ITV France method [75]: 50 g of crushed berry mash was homogenised in 85 mL of 0.1% HCl (diluted from 37% hydrochloric acid, Carl Roth GmbH + Co. KG, Karlsruhe, Germany) with 15 mL of 95% vol. ethanol (diluted from absolute ethanol, Merck KGaA, Darmstadt, Germany) and macerated for 1 h at 20 °C (1 min homogenisation every 15 min). After filtration, the samples were filtered and diluted 1:100 with distilled water for optical density measurement at 280 nm (total polyphenolic index, AU Kg⁻¹) and 1:20 with 0.1% HCl for optical density measurement at 520 nm (potential anthocyanins, mg Kg⁻¹), using a UV-VIS double-beam spectrophotometer (Specord 250, Analytik Jena AG, Jena, Germany).

2.3.6. Changes in Berry Weight and Sugar Content During Ripening

In addition to the harvest maturity fruit quality parameters (Section 2.3.5), an exploratory time-resolved dataset of berry weight and total soluble solids was collected from véraison to post-harvest maturity for both cultivars (Supplementary Figure S2). At each vine, berries were collected in sub-groups of 10–15, sampled twice from each of the sun-exposed and shaded sides of the canopy, and pooled across three vines per cultivar to form a single composite sample of approximately 120–150 berries per date; these values are therefore presented as descriptive field observations rather than as biological replicate means, consistent with weight-based berry sugar content estimation as previously validated against HPLC-measured sugar [76]. Berries were weighed and total soluble solids (°Brix) determined by refractometry. Sugar content per berry (g berry⁻¹) was calculated by converting °Brix to sugar concentration (g Kg⁻¹) using OIV reference tables [77], then applying this concentration to mean single-berry weight (100-berry weight ÷ 100), following the active-versus-passive ripening framework of Casciato and Keller [78].

2.4. Experimental Design and Statistical Analysis

The experiment was conducted on two grapevine cultivars (C), with three biological replicates (vines) per cultivar included in the statistical analysis. Physiological measurements were performed longitudinally at three phenophases (flowering, berry growth, and véraison) for gas exchange parameters (A, E, g_s, C_i and WUE) and at four phenophases (including maturity) for biochemical and structural indicators (Chl a, Chl b, C+X, dm, ratio Chl/(C+X) and Chl a/b).

Statistical analyses were performed using IBM SPSS Statistics software (version 26.0, IBM Corp., Armonk, NY, USA). The significance level was set at α = 0.05, and differences were considered statistically significant when p ≤ 0.05.

The experimental data were analysed using a Linear Mixed-Effects Model (LMM) to account for the dependency of repeated measurements within the same biological units (vines). Grape cultivar (C), phenophase (P), and their interaction (C × P) were included as fixed effects. Phenophase (P) was modelled as a repeated factor within each vine. Model was fitted using a first-order autoregressive covariance (AR1) structure to account for the temporal autocorrelation between successive phenophases. The Satterthwaite approximation was employed to estimate the degrees of freedom and adjust for potential heteroscedasticity across cultivars and growth stages. The model parameters were estimated using Restricted Maximum Likelihood (REML). Post hoc pairwise comparisons of estimated marginal means (EMMs) were performed using the Bonferroni adjustment to maintain the family-wise error rate.

The multivariate statistical techniques were performed using OriginPro 2024 (v10.1, OriginLab Corporation, Northampton, MA, USA). Principal Component Analysis (PCA) was used to lower the dataset’s dimensionality and visualise the spatial separation between cultivars and phenophases based on physiological and biochemical responses of Fetească neagră and Șarba grape cultivars.

In order to ensure that all physiological and biochemical parameters, regardless of their initial units or magnitudes, contributed equally to the overall variance, the PCA was calculated using a correlation matrix on Z-score standardised data. Principal components with eigenvalues > 1 were retained for analysis, according to the Kaiser criterion. A heatmap virtual matrix was constructed on averaged standardised Z-scores of biological replicates to highlight relative changes across the phenophases for each variable. This approach graphically depicts the intensity and direction of physiological and biochemical changes across phenophases. To verify the observed grouping patterns, a Hierarchical Cluster Analysis (HCA) was performed (presented in Supplementary Figure S1). Using the maximum distance between clusters, the furthest neighbour (complete linkage) method was used for clustering, and the squared Euclidean distance was used to calculate dissimilarities.

3. Results

3.1. Climatic Characterisation of the 2024 Growing Season

Table 1. Climatic indicators of 2024 compared to the multiannual average (1991–2020).

Climatic Indicator	Multiannual Average (1991–2020)	2024	Difference (±)
Average annual temperature, °C	10.98	14.70	+3.72
Average temperature in the period IV–VIII, °C	18.93	22.21	+3.28
Average temperature in the growing season, °C (IV–X)	17.46	20.62	+3.16
Average temperature in summer, °C (VI–VIII)	22.15	26.30	+4.15
Average annual minimum temperature, °C	5.24	7.70	+2.46
Average annual maximum temperature, °C	17.6	21.10	+3.50
Number of hot days (30 °C < Tmax ≤ 35 °C)	46	52	+6
Number of very hot days (Tmax > 35 °C)	10	38	+28
Annual total precipitation, mm	633	583	−50
Total precipitation in the period IV–VIII, mm	315.3	257.5	−57.8
Total precipitation in the growing season (IV–X), mm	430	339	−91
Total precipitation in summer (VI–VIII), mm	193	180	−13
Hydrothermal coefficient (HC) [67]	1.1	0.75	−0.35
Huglin index (HI) [68]	3163	4138	+975
Winkler index (WI) [69]	1762	2273	+511
De Martonne Aridity Index (DMI) [70]	30.3	23.6	−6.70
Cool night index IX (CNI) [71]	10.64	13.5	+2.86
* Cool night index VIII (CNI) [71]	15.13	17.13	+2.00

* CNI calculated for August 2024 (VIII) to reflect the phenologically advanced harvest under the exceptional thermal conditions of the study year (Huglin Index: 4138; +4.15 °C above the 1991–2020 baseline). Following Tonietto and Carbonneau [71], the CNI conventionally uses September (IX) as the reference month for standard growing seasons.

presents the main climatic variables and bioclimatic indices recorded in 2024 compared to the multiannual reference period (1991–2020).

The year 2024 was markedly warmer than the multiannual average (1991–2020) across all thermal indicators. The largest anomalies were recorded for mean summer temperature (June–August: +4.15 °C), mean annual temperature (+3.72 °C), and mean annual maximum temperature (+3.50 °C) (Table 1).

Aridity indices confirmed semi-arid conditions: the De Martonne Index (DMI) fell to 23.6 (vs. 30.3 for 1991–2020) and the Hydrothermal Coefficient dropped to 0.75 (vs. 1.1). The number of extremely hot days (Tmax > 35 °C) reached 38, approximately 3.8 times the historical average of 10, underscoring the exceptional intensity of heat stress during the growing season.

The Cool Night Index (CNI) was calculated for August rather than the standard September [71], reflecting the phenologically advanced harvest under the exceptional thermal conditions of 2024 (Huglin Index: 4138; mean summer temperature +4.15 °C above the 1991–2020 baseline). This adjustment is consistent with the original definition of CNI as an indicator of night temperature conditions during the final maturation period, rather than a fixed calendar month.

Precipitation declined across all measured intervals, by 91 mm during the growing season and 50 mm annually, compounding the thermal stress and contributing to progressive drought-induced berry water loss observed across both cultivars by harvest maturity (Figure 2).

Both the Hydrothermal Coefficient (0.75) and the De Martonne Index (23.6) confirm the semi-arid character of the 2024 season, consistent with conditions under which supplemental irrigation is typically regarded as beneficial for wine grape cultivation [67,70]. More striking still, the Huglin Index climbed to 4138, surpassing both the multiannual average (3163) and the ceiling of the very warm climate class (HI+ zone 3), and situating the site in a thermal regime now characteristic of Málaga, Spain, or Marsala, Italy [58]. The Winkler Index and Cool Night Index told the same story, each exceeding their multiannual averages and confirming that 2024 was an anomaly across every measured thermal dimension.

3.2. Phenophases in 2024

Table 2 presents the mean day of year (DOY) and useful heat balance (UHB) at each of the four main phenophases recorded at the USAMV Bucharest ampelographic collection in 2024. Differences in phenological timing between cultivars are attributable to genetic factors, climatic conditions, soil characteristics, and viticultural practices [20,21].

Table 2. Mean day of year (DOY) and useful heat balance (UHB, °C) for phenophases, USAMV Bucharest ampelographic collection, 2024.

Grape Cultivar	Budburst (BBCH 008)		Flowering (BBCH 605)		Véraison (BBCH 801)		Harvest Maturity (BBCH 809)
	Date (DOY)	UHB * (°C)	Date (DOY)	UHB * (°C)	Date (DOY)	UHB * (°C)	Date (DOY)	UHB * (°C)
Șarba	96 (5 April)	18.5	146 (25 May)	239.5	220 (7 August)	1086.5	239 (26 August)	308.0
Fetească neagră	96 (5 April)	18.5	146 (25 May)	239.5	204 (22 July)	866.0	235 (22 August)	461.5

* UHB: useful heat balance, calculated as UHB = Σ(Tmean,i − 10) for days where Tmean,i > 10 °C. Each phenophase date (DOY) is the single day of 50% occurrence (Section 2.3.1); it does not represent a duration. The UHB value shown is the cumulative sum of degree-days from the end of the preceding phenophase up to and including the listed date—i.e., the heat accumulated specifically during the interval leading into that phenophase.

Table 2. Mean day of year (DOY) and useful heat balance (UHB, °C) for phenophases, USAMV Bucharest ampelographic collection, 2024.

Grape Cultivar	Budburst (BBCH 008)		Flowering (BBCH 605)		Véraison (BBCH 801)		Harvest Maturity (BBCH 809)
	Date (DOY)	UHB * (°C)	Date (DOY)	UHB * (°C)	Date (DOY)	UHB * (°C)	Date (DOY)	UHB * (°C)
Șarba	96 (5 April)	18.5	146 (25 May)	239.5	220 (7 August)	1086.5	239 (26 August)	308.0
Fetească neagră	96 (5 April)	18.5	146 (25 May)	239.5	204 (22 July)	866.0	235 (22 August)	461.5

* UHB: useful heat balance, calculated as UHB = Σ(Tmean,i − 10) for days where Tmean,i > 10 °C. Each phenophase date (DOY) is the single day of 50% occurrence (Section 2.3.1); it does not represent a duration. The UHB value shown is the cumulative sum of degree-days from the end of the preceding phenophase up to and including the listed date—i.e., the heat accumulated specifically during the interval leading into that phenophase.

Rising temperatures exert their most immediate effect on the advancement of grapevine phenophases. Differences in phenological timing and in the intervals between stages reflect the combined influence of genetic factors, climate, soil conditions, and viticultural practices [21].

Table 2 presents the mean day of year (DOY) and cumulative useful heat balance (UHB) for the four phenophases—budburst, flowering, véraison, and harvest maturity—recorded in 2024. As expected under the exceptional thermal conditions of the study year, the greatest inter-cultivar divergence occurred at véraison and harvest maturity, the phenophases most sensitive to accumulated heat and water status.

Both cultivars shared identical budburst and flowering dates, reflecting the shared early-season thermal environment. Véraison (50% berry softening) was recorded on DOY 204 (22 July) for Fetească neagră and DOY 220 (7 August) for Șarba—a difference of 16 days consistent with the earlier-ripening character of the former. Harvest maturity was reached on 22 August for Fetească neagră and 26 August for Șarba, approximately nine days earlier than the 2015–2019 reference period for Șarba at the same site [21], a direct consequence of the exceptional thermal accumulation of 2024.

3.3. Effects of Cultivar, Phenophase,, and Their Interaction on Physiological Parameters

The LMM (Linear Mixed Model) statistical analysis (

Table 3. Linear Mixed Model (LMM) results for the impact of grape cultivar (C) and phenophase (P) on physiological indicators of grapevines.

Physiological Indicators	Grape Cultivar (C)		Phenophase (P)		Interaction C × P
	F-Statistic	p-Value	F-Statistic	p-Value	F-Statistic	p-Value
A—Net photosynthesis (μmol CO2 m−2 s−1)	7.70 (df1 = 1, df2 = 5.73)	0.034	28.47 (df1 = 2, df2 = 9.00)	0.000	0.42 (df1 = 2, df2 = 9.00)	0.671
E—Transpiration rate (mmol H2O m−2 s−1)	99.73 (df1 = 1, df2 = 6.57)	0.000	78.01 (df1 = 2, df2 = 9.46)	0.000	7.39 (df1 = 2, df2 = 9.46)	0.012
gs—Stomatal conductance (mol H2O m−2 s−1)	36.48 (df1 = 1, df2 = 7.22)	0.000	47.34 (df1 = 2, df2 = 9.89)	0.000	7.53 (df1 = 2, df2 = 9.89)	0.010
Ci—Intercellular CO2 (µmol CO2 mol −1)	1.31 (df1 = 1, df2 = 5.60)	0.299	9.10 (df1 = 2, df2 = 8.66)	0.007	2.46 (df1 = 2, df2 = 8.66)	0.143
WUE—Water use efficiency (A/E)	5.95 (df1 = 1, df2 = 4.10)	0.070	19.72 (df1 = 2, df2 = 7.64)	0.001	18.36 (df1 = 2, df2 = 7.64)	0.001
Chl a—Chlorophyll a (mg 100 g−1 FW)	5.20 (df1 = 1, df2 = 6.25)	0.061	4.22 (df1 = 3, df2 = 11.85)	0.030	1.03 (df1 = 3, df2 = 11.85)	0.416
Chl b—Chlorophyll b (mg 100 g−1 FW)	18.37 (df1 = 1, df2 = 8.42)	0.002	3.48 (df1 = 3, df2 = 12.46)	0.049	2.55 (df1 = 3, df2 = 12.46)	0.103
C+X—Carotenoids (mg 100 g−1 FW)	23.61 (df1 = 1, df2 = 5.02)	0.005	30.05 (df1 = 3, df2 = 11.11)	0.000	19.02 (df1 = 3, df2 = 11.11)	0.000
dm—Dry matter (% w/w)	4.15 (df1 = 1, df2 = 7.95)	0.076	81.27 (df1 = 3, df2 = 10.93)	0.000	2.55 (df1 = 3, df2 = 10.93)	0.109
Chl/(C+X)—ratio total chlorophyll/total carotenoids	35.92 (df1 = 1, df2 = 6.78)	0.001	12.98 (df1 = 3, df2 = 11.62)	0.001	10.72 (df1 = 3, df2 = 11.62)	0.001
Chl a/b—ratio Chlorophyll a/Chlorophyll b	11.70 (df1 = 1, df2 = 8.31)	0.009	1.84 (df1 = 3, df2 = 12.44)	0.193	1.38 (df1 = 3, df2 = 12.44)	0.294

) revealed that phenophase (P) showed a statistically significant effect for most, but not all, physiological indicators examined, with the exception of the Chl a/b ratio, for which the phenophase effect was not significant (p = 0.193). Grape cultivar (C) showed significant differences for net photosynthesis (A, p = 0.034), transpiration rate (E, p < 0.001), stomatal conductance (g_s, p < 0.001), chlorophyll b (Chl b, p = 0.002), and total carotenoids (C+X, p = 0.005), the Chl/(C+X) ratio (p = 0.001), and the Chl a/b ratio (p = 0.009). Significant C×P interaction effects were detected for E (p = 0.012), g_s (p = 0.010), WUE (p = 0.001), and C+X (p < 0.001), indicating that the two cultivars respond differently to environmental conditions as they progress through phenophases.

Pairwise comparisons using LMM Bonferroni post hoc tests (

Table 4. Pairwise comparisons using Linear Mixed Model (LMM) post hoc comparisons using Bonferroni corrections.

Physiological Indicators	Grape Cultivar (C)	Phenophase (P)
		Flowering	Berry Growth	Véraison	Maturity
A—Net photosynthesis (μmol CO2 m−2 s−1)	S	6.8 ± 0.87 aA	1.81 ± 0.87 bA	3.29 ± 0.87 bB	-
	FN	8.81 ± 0.87 aA	4.2 ± 0.87 bA	6.53 ± 0.87 abA	-
E—Transpiration rate (mmol H2O m−2 s−1)	S	4.93 ± 0.36 aB	1.25 ± 0.36 bB	1.04 ± 0.36 bB	-
	FN	7.75 ± 0.36 aA	2.76 ± 0.36 cA	5.28 ± 0.36 bA	-
gs—Stomatal conductance (mol H2O m−2 s−1)	S	0.13 ± 0.02 aB	0.02 ± 0.02 bA	0.03 ± 0.02 bB	-
	FN	0.32 ± 0.02 aA	0.05 ± 0.02 bA	0.12 ± 0.02 bA	-
Ci—Intercellular CO2 (µmol CO2 mol−1)	S	311.93 ± 17.82 aA	276.6 ± 17.82 aA	236.4 ± 17.82 aA	-
	FN	347.13 ± 17.82 aA	246.33 ± 17.82 bA	278.07 ± 17.82 abA	-
WUE—Water use efficiency (A/E)	S	1.36 ± 0.26 bA	1.59 ± 0.26 bA	3.26 ± 0.26 aA	-
	FN	1.18 ± 0.26 aA	1.57 ± 0.26 aA	1.25 ± 0.26 aB	-
Chl a—Chlorophyll a (mg 100 g−1 FW)	S	101.48 ± 4.44 abA	99.92 ± 4.44 bA	119.14 ± 4.44 aA	109.86 ± 4.44 abA
	FN	92.21 ± 4.44 aA	100.02 ± 4.44 aA	104.46 ± 4.44 aB	99.08 ± 4.44 aA
Chl b—Chlorophyll b (mg 100 g−1 FW)	S	58.12 ± 6.91 aA	63.09 ± 6.91 aA	83.44 ± 6.91 aA	89.57 ± 6.91 aA
	FN	46.09 ± 6.91 aA	59.13 ± 6.91 aA	60.4 ± 6.91 aB	49.87 ± 6.91 aB
C+X—Carotenoids (mg 100 g−1 FW)	S	26.11 ± 0.82 aA	26.24 ± 0.82 aA	20.61 ± 0.82 bB	16.08 ± 0.82 cB
	FN	25.34 ± 0.82 abA	26.38 ± 0.82 abA	27.7 ± 0.82 aA	24.62 ± 0.82 bA
dm—Dry matter (% w/w)	S	29.43 ± 0.82 bA	31.05 ± 0.82 bA	36.17 ± 0.82 aA	38.18 ± 0.82 aA
	FN	25.95 ± 0.82 dB	30.64 ± 0.82 cA	34.59 ± 0.82 bA	39.85 ± 0.82 aA
Chl/(C+X)—ratio total chlorophyll/total carotenoids	S	6.12 ± 0.62 bA	6.26 ± 0.62 bA	9.89 ± 0.62 aA	12.36 ± 0.62 aA
	FN	5.51 ± 0.62 aA	6.05 ± 0.62 aA	5.96 ± 0.62 aB	6.05 ± 0.62 aB
Chl a/b—ratio Chlorophyll a/Chlorophyll b	S	1.77 ± 0.16 aA	1.6 ± 0.16 aA	1.43 ± 0.16 aA	1.27 ± 0.16 aB
	FN	2.09 ± 0.16 aA	1.71 ± 0.16 aA	1.74 ± 0.16 aA	2.01 ± 0.16 aA

Grape cultivar: S—Șarba and FN—Fetească neagră; lowercase letters show differences between different phenophases (P); uppercase letters show differences between grape cultivars (C); standard errors were derived from the estimated marginal means of the grape cultivar (C) × phenophase (P) interaction to reflect the specific error variance associated with each phenophase.

) revealed distinct patterns. At véraison, Fetească neagră maintained a transpiration rate (E) of 5.28 mmol H₂O m⁻² s⁻¹, nearly five times higher than Șarba (1.04 mmol H₂O m⁻² s⁻¹) and, exhibited significantly higher net photosynthesis (A: 6.53 vs. 3.29 μmol CO₂ m⁻² s⁻¹). Despite consuming CO₂ at nearly double the rate, Fetească neagră maintained stable intercellular CO₂ concentrations (C_i, p = 0.299, Table 3) through significantly higher stomatal conductance (g_s = 0.12 vs. 0.03 mol H₂O m⁻² s⁻¹, Table 4). In contrast, Șarba exhibited a pronounced water-conserving stomatal response at véraison, with stomatal closure limiting transpiration to 1.04 mmol H₂O m⁻² s⁻¹ and net photosynthesis to 3.29 μmol CO₂ m⁻² s⁻¹. The significant C × P interaction for g_s (p = 0.010) and E (p = 0.012, Table 3) statistically confirms that Șarba undergoes more severe stomatal-driven downregulation during ripening compared to Fetească neagră.

The LMM (Table 3) revealed a highly significant C × P interaction for the Chl/(C+X) ratio (p = 0.001), indicating that the two cultivars manage their photosynthetic pigment composition differently as they progress through the season. Fetească neagră maintained a stable Chl/(C+X) ratio throughout all phenophases, whereas Șarba significantly increased this ratio at véraison (9.89 ± 0.62 vs. 5.96 ± 0.62, Table 4), coinciding with a significant decrease in carotenoid content (C+X: 20.61 ± 0.82 mg 100 g⁻¹ FW). Conversely, Fetească neagră preserved carotenoid content at véraison (C+X: 27.7 ± 0.82 mg 100 g⁻¹ FW). Leaf dry matter content (dm) showed significant phenophase effects (p < 0.001), with Șarba maintaining consistently higher dm% values across all phenophases.

3.4. Multivariate Analysis: Divergent Physiological Behaviour

The PCA biplot (Figure 3) captured 78.42% of total variance (PC1: 62.17%; PC2: 16.25%). As shown in Supplementary Table S1, PC1 was defined by high positive loadings for E (0.327), g_s (0.297), and A (0.241), and negative loadings for Chl a (−0.311), Chl b (−0.356), and dm (−0.295). PC2 was characterised by a strong negative loading for C+X (−0.478) and positive loadings for WUE (0.288) and A (0.414).

Data points represent vine replicates, categorised by cultivar (e.g., FN for Fetească neagră and S for Șarba) and phenophase (f: flowering; bg: berry growth; v: véraison). Loadings (blue vectors): the direction and length of the vectors indicate the loading of each variable (A, E, g_s, C_i, WUE, Chl a, Chl b, Chl a/b, C+X, Chl/(C+X), dm) on the principal components. The light blue and light red dashed arrows indicate the divergent water-conserving response (Șarba) and gas exchange sustaining response (Fetească neagră), respectively. The central grey circle includes the clustered samples with average metabolic activity, primarily during the berry growth phase (Sb and FNb) and the flowering stage of Șarba (Sf) or the véraison stage of the resilient variety Fetească neagră (FNv).

Véraison was the phenophase at which the water-conserving and gas exchange sustaining strategies became most clearly differentiated (Supplementary Table S1). Șarba at véraison (Sv) occupied the upper-left quadrant, characterised by peak WUE and elevated Chl/(C+X), consistent with a pronounced metabolic downregulation and stomatal conservation of water. Fetească neagră at véraison (FNv), by contrast, retained an intermediate metabolic profile, sustaining moderate gas exchange rates rather than closing the stomata, while Fetească neagră at flowering (FNf) anchored the gas exchange sustaining extreme of the biplot, associated with the highest A, E, and g_s values. Both cultivars converged in a central intermediate cluster during berry growth, reflecting a period of comparable physiological activity that precedes their sharp divergence at véraison.

The Pearson correlation matrix (Supplementary Table S2) was consistent with stomatal regulation as being the principal physiological driver of water status in both cultivars, evidenced by the near-perfect correlation between g_s and E (r = 0.933). Strong positive correlations among A, E, and g_s (r > 0.74) were reflected in their high positive PC1 loadings, confirming that gas exchange parameters co-vary as an integrated stomatal response. The strong negative correlation between Chl/(C+X) and C+X (r = −0.905) quantifies the inverse relationship between carotenoid accumulation and the chlorophyll-to-carotenoid ratio, supporting the photoprotective role of carotenoids under thermal stress. The negative WUE–C_i correlation (r = −0.597) is consistent with the water-saving strategy of Șarba operating at the direct cost of internal CO₂ availability, while the strong positive WUE–Chl/(C+X) correlation (r = 0.855) is consistent with a link between water conservation and the maintenance of chlorophyll integrity relative to carotenoids.

The colour gradient represents relative deviations from the mean (red: high; blue: low). Light grey areas indicate missing values for the berry maturity phase (Sm, FNm). Distinct clustering reveals the divergent water-conserving strategy of Șarba at véraison (Sv), characterised by peak WUE, and the gas exchange sustaining response of Fetească neagră (FNv), which sustains intermediate metabolic activity. Cultivar/phenophase codes: S = Șarba, FN = Fetească neagră; f = flowering, b = berry growth, v = véraison, m = harvest maturity (e.g., Sv = Șarba at véraison, FNm = Fetească neagră at harvest maturity).

The Z-score standardised heatmap (Figure 4) confirmed the divergence in resource management strategies: only Fetească neagră (FNv) maintained positive metabolic scores for carbon assimilation (A) at véraison, while Șarba (Sv) showed the highest WUE values alongside declining carotenoid-based photoprotection.

The HCA dendrogram (Supplementary Figure S1), constructed using Squared Euclidean Distance, resolved three distinct clusters: a water-conserving Sv cluster (blue branch), a gas exchange sustaining FNf cluster (red branch), and a central intermediate metabolic group comprising berry growth samples of both cultivars.

3.5. Grape Quantitative and Qualitative Parameters at Harvest Maturity

At harvest, Fetească neagră reached a higher sugar concentration (27.6 °Brix) than Șarba (22.0 °Brix), despite smaller berries (94 vs. 154 g per 100 berries) and a smaller cluster weight (205 vs. 266 g) (

Table 5. Grape quantitative and qualitative parameters at harvest maturity for Șarba and Fetească neagră, USAMV Bucharest experimental vineyard, 2024.

Variety	Cluster Weight (g)	Weight of 100 Berries (g)	Sugar Content (°Brix)	Titratable Acidity (g L−1 Tartaric Acid)	pH	The Polyphenolic Potential of Grapes		Anthocyanins (mg L−1)	Total Anthocyanin Potential (mg Kg−1)
						OD 280 nm	Polyphenolic Index
Șarba	266	154	22.0	3.64	3.7	0.106	31.8	-	-
Fetească neagră	205	94	27.6	4.13	3.6	0.243	72.9	270.9	811.3

). The polyphenolic index was more than double in Fetească neagră (72.9 vs. 31.8 AU Kg⁻¹), with anthocyanins reaching 270.9 mg L⁻¹ and a total anthocyanin potential of 811.3 mg Kg⁻¹.

Supplementary Figure S2 extends this single harvest maturity measurement with a dataset capturing the dynamics of berry weight and total soluble solids from véraison to post-harvest maturity, discussed further in Section 4.2.

4. Discussion

Two Romanian autochthonous cultivars, Șarba and Fetească neagră, were studied under identical field conditions during a climatically exceptional season, and they responded in fundamentally different ways. Across all analytical approaches applied—LMM, PCA, and Pearson correlation—Șarba’s responses were consistent with near-isohydric behaviour and Fetească neagră’s with anisohydric behaviour, with the semi-arid conditions of 2024 (DMI: 23.6; HC: 0.75) providing the environmental severity needed to make that contrast fully legible. The following subsections interpret this divergence: Section 4.1 addresses baseline climatic and phenological data. Section 4.2 and Section 4.3 evaluate our empirical gas exchange and pigment measurements alongside candidate physiological mechanisms drawn from the literature—which, while consistent with our findings, are not directly tested by the present dataset. Section 4.4 translates these physiological insights into cultivar-specific management implications, and Section 4.5 addresses study limitations explicitly.

4.1. Semi-Arid Conditions and Phenological Advancement

The exceptional climatic conditions of 2024 brought 3.8-fold increase in very hot days (Tmax > 35 °C) and a growing season precipitation deficit of 91 mm, which are consistent with the documented progressive aridification of Romanian viticultural regions [50,51,52,53,54,55,56]. The phenological consequence was immediate. Harvest advanced approximately nine days earlier in Șarba than in the 2015–2019 reference period at the same site [21], making the August CNI calculation not an adjustment but a physiological necessity, the vine had simply moved past September before that month arrived.

4.2. Divergent Stomatal Regulation Strategies

Stomatal density was numerically higher in Șarba, but the absence of statistical significance (p > 0.05) ruled out anatomy as the driver of divergence. What differentiated the two cultivars was behaviour, specifically, how each regulated stomatal aperture as the season progressed toward véraison. This is consistent with the functional framework established for grapevine drought responses [24,25].

Fetească neagră kept its stomata open in a manner consistent with a progressive decline in leaf water potential, sustaining high CO₂ assimilation and transpirational cooling throughout the season, an anisohydric behaviour also documented in wild Vitis vinifera subsp. sylvestris [79]. Its preserved carotenoid content at véraison appears to provide the photoprotective capacity needed to maintain this metabolic rate under extreme thermal load [31,35]. In the anisohydric Sangiovese, Dal Santo et al. [31] demonstrated that ROS-scavenging enzymes and abiotic stress genes are induced earlier and more strongly than in isohydric cultivars, a molecular defence profile consistent with the sustained photosynthetic activity and cell viability observed in Fetească neagră under the 2024 drought.

Șarba’s stomatal conductance declined progressively across the season, from 0.13 mol H₂O m⁻² s⁻¹ at flowering to 0.02 mol H₂O m⁻² s⁻¹ at berry growth and 0.03 mol H₂O m⁻² s⁻¹ at véraison—an approximately 77% reduction from flowering to véraison—consistent with the near-isohydric stomatal behaviour documented in Muscat-type genotypes [80].

The confirmed Muscat-type ancestry of Șarba [21,62] may underpin this constitutive water conservation tendency genetically. Early stomatal closure in Șarba suppresses net photosynthesis, reducing the carbon supply available to the plastidial MEP (methylerythritol 4-phosphate) pathway from which GPP (Geranyl PyroPhosphate), the obligate precursor for monoterpene biosynthesis, is derived [81]. Under moderate water deficit, the MEP pathway can be upregulated at the transcriptional level via drought-responsive and ABA-mediated regulatory elements, increasing monoterpene accumulation in white grape berries [82]. However, the severe stomatal shutdown observed here (A = 3.29 μmol CO₂ m⁻²s⁻¹ at véraison; g_s = 0.03 mol H₂O m⁻² s⁻¹) risks depleting the carbon substrate pool beyond the threshold at which this transcriptional induction can be sustained, potentially compromising the aromatic typicity that defines Șarba as a semi-aromatic variety.

Pou et al. [83], working with two-year-old potted vines under moderate controlled stress with all bunches removed, found that anisohydric Chardonnay achieved higher instantaneous WUE (A/E) than isohydric cultivars, partly through reduced leaf-to-air vapour pressure deficit maintained by sustained transpirational cooling. The present study was conducted on mature field vines over 20 years old, carrying a full crop load under semi-arid conditions that differ fundamentally from controlled pot experiments. Under these field conditions, Șarba’s own WUE (A/E) rose sharply as the season progressed—from 1.36 at flowering and 1.59 at berry growth to 3.26 at véraison, a 2.4-fold increase, reflecting the dominance of stomatal limitation over atmospheric demand once drought exceeds the moderate threshold at which anisohydric advantages operate. Fetească neagră’s WUE, by contrast, remained comparatively stable across the same period (1.18, 1.57, and 1.25 at flowering, berry growth, and véraison, respectively). Despite its lower net photosynthesis at véraison, Șarba did not show correspondingly higher sugar accumulation at harvest (22.0 vs 27.6 °Brix in Fetească neagră, Table 5), consistent with the reduced carbon assimilation throughout the ripening period.

Under prolonged drought, the anisohydric behaviour of Fetească neagră is presumed to operate with narrower hydraulic safety margins, by analogy with other anisohydric varieties in which this has been directly measured, and therefore may carry significant physiological risk, despite its sustained capacity for carbon assimilation [84]. Anisohydric behaviour is generally associated with drought tolerance [84,85,86] through open stomata which sustain carbon gain and osmotic adjustment, and a similar ROS-scavenging response, as described in anisohydric Sangiovese by Dal Santo et al. [31], could plausibly contribute to maintaining cellular defence capacity under gradual water loss, though this was not directly measured in the present study. Yet when deficit extends beyond the vine’s hydraulic safety margins, this strategy reverses from an advantage into a liability.

Premature turgidity loss and senescence of basal leaves have been visually observed in Fetească neagră vineyards across Romania in recent years, but to our knowledge this study provides the first formal documentation of this phenomenon for the cultivar (Figure 5).

The spatial pattern of stress-induced leaf senescence further reflects this mechanistic divergence. In Fetească neagră, basal leaf senescence is consistent with arising from the hydraulic gradient amplified by xylem backflow, a proposed mechanism whereby leaves at the shoot base—where water potential could be expected to be lowest given the hydraulic gradient—would lose turgor first, producing a spatially localised response that progresses acropetally. In Șarba, by contrast, prolonged near-isohydric stomatal closure risks whole-canopy carbon starvation, expressed—when drought is sufficiently extended—as diffuse general yellowing rather than basal-specific senescence, consistent with the pattern reported in near-isohydric varieties with tight stomatal control under drought [31,87]. Although Șarba’s near-isohydric protection delays this response relative to Fetească neagră, the low net photosynthesis recorded at véraison (A = 3.29 μmol CO₂ m⁻² s⁻¹) indicates that the carbon balance was already under strain during the 2024 season, with whole-canopy yellowing as the probable next step under further or repeated drought.

The cellular mechanism underlying these contrasting responses matters for wine quality. In Fetească neagră, progressive hydraulic water loss driven by amplified xylem backflow is presumed to occur against a background of sustained metabolic activity. Phloem loading is assumed to continue as cells lose water gradually, while membrane integrity is maintained through osmotic adjustment. Sugars are likewise presumed to concentrate passively, rather than through active import arrest. This proposed mechanism is broadly consistent with a supplementary, exploratory analysis of berry weight and soluble solids dynamics (Supplementary Figure S2). Through the middle of the ripening period, Fetească neagră’s estimated sugar content (g berry⁻¹) rises alongside increasing berry mass—consistent with continued active sugar delivery. In the final sampling interval, however, berry mass declines and estimated sugar content (g berry⁻¹) plateaus and slightly declines, even as °Brix continues to rise—a pattern consistent with passive concentration becoming the dominant process only late in ripening. This two-phase pattern is directly analogous to that recently documented for active-versus-passive sugar concentration in Cabernet Sauvignon, Merlot, and Syrah [78]. As this dataset consists of single composite samples without biological replication (Section 2.3.6), it is presented as corroborating context rather than a formal test of the proposed mechanism. In Șarba, the acute berry shrivelling observed in Figure 2 reflects a different pathway. Once shoot water potential is presumed to drop below the turgor loss point despite near-isohydric protection, the cascade that follows (loss of membrane integrity, brush-region desiccation, and arrest of sugar accumulation) is consistent with the hallmark presentation of berry shrivel disorder [44,45], though not independently confirmed by the present measurements. In practical terms, this contrast suggests that Fetească neagră’s living-cell dehydration, if controlled, could yield desirable phenolic concentration, whereas Șarba’s dead-cell shrivelling would be expected to produce raisined, oxidised fruit. Anthocyanin synthesis and tannin polymerisation are additionally inhibited above 30 °C [10,88,89], while stomatal conductance also declines significantly when air temperature exceeds the optimum photosynthetic range in grapevine [90], compounding quality and physiological risk when heat and drought co-occur, as they did throughout the 2024 ripening period (Table 1).

4.3. Photoprotective Pigment Response

The divergent Chl/(C+X) trajectories reveal the two cultivars managing their photosynthetic apparatus in opposite directions as véraison approaches. Șarba’s own carotenoid content declined progressively across the season, from 26.24 mg 100 g⁻¹ FW at berry growth to 20.61 mg 100 g⁻¹ FW at véraison and 16.08 mg 100 g⁻¹ FW at harvest, a pattern consistent with isohydric stomatal closure: when stomata shut and mesophyll irradiance falls, the demand for carotenoid-based photoprotection is expected to diminish, and the elevated Chl/(C+X) ratio (9.89 ± 0.62) is consistent with reduced carotenoid deployment rather than increased chlorophyll. Fetească neagră’s carotenoid content, by contrast, changed comparatively little across the same interval (26.38 mg 100 g⁻¹ FW at berry growth to 27.7 mg 100 g⁻¹ FW at véraison, declining more moderately to 24.62 mg 100 g⁻¹ FW at harvest), consistent with a photoprotective buffer sustaining photosynthesis under high irradiance, in line with the photoprotective function of this ratio described by Villalobos-González et al. [35].

The significant cultivar effect on Chl a/b (p = 0.009), independent of phenophase or their interaction, points to a sustained difference in chlorophyll allocation within the photosynthetic apparatus, with Fetească neagră maintaining a higher Chl a/b ratio than Șarba throughout the season. Since Chl b is concentrated almost exclusively in the LHCII peripheral antenna of PSII, and chlorophyll b reductase preferentially degrades Chl b-associated LHCII under high-light exposure [91], a higher Chl a/b ratio is generally indicative of a reduced LHCII pool relative to the photosynthetic core. The sustained high transpiration rates (E = 5.28 mmol H₂O m⁻² s⁻¹ at véraison) and open stomata characteristic of anisohydric behaviour—consistent with a progressive decline in leaf water potential—expose Fetească neagră’s leaves to greater cumulative photoinhibitory stress, consistent with the reduced chlorophyll stability reported under water deficit in other grapevine cultivars [92]. This pigment-level signature aligns with and, may help explain, the basal leaf senescence observed in Fetească neagră (Figure 5). In Șarba, by contrast, early stomatal closure limits photoinhibitory exposure, allowing the Chl b-rich LHCII pool to be maintained and even increase toward maturity (Table 4), consistent with its lower and more stable Chl a/b ratio.

The near-perfect g_s–E correlation (r = 0.933) and the strong negative WUE–C_i correlation (r = −0.597) add quantitative precision to this picture: stomatal regulation is consistent with being the principal driver of water status in both cultivars, and the water conservation that defines Șarba’s véraison strategy carries a measurable cost in internal CO₂ availability, with direct implications for late-season carbon assimilation and aromatic precursor biosynthesis.

It should also be noted that the extreme temperatures of the 2024 season (Table 1) may have independently contributed to the chlorophyll changes observed (Table 4), as high temperatures have been shown to reduce leaf chlorophyll content in grapevine regardless of water status [90]—a temperature-driven component that could not be fully separated from the drought effects in this study.

4.4. Cultivar-Specific Management Implications

The interventions outlined below represent technological possibilities grounded in evidence from other grapevine cultivars or, in some cases, other crop species; none have yet been tested in Șarba or Fetească neagră, and their efficacy under the semi-arid conditions described here remains to be established.

4.4.1. Șarba

Regulated deficit irrigation (RDI) during the véraison-ripening period would be expected to help prevent the complete stomatal closure characteristic of Șarba’s water-conserving response, potentially maintaining evaporative cooling, sustaining carbon assimilation, and preserving the GPP precursor pool necessary for terpene biosynthesis and aromatic typicity [81]. In Gewürztraminer, an aromatic Muscat-family cultivar comparable in typicity to Șarba, late-season RDI increased key free terpenes at harvest without affecting yield, while terpene gene expression remained unchanged, indicating that the gain reflects metabolic substrate availability rather than transcriptional induction [93]. The lower sugar and polyphenolic index recorded in Șarba at harvest (22.0 °Brix; 31.8 AU Kg⁻¹, Table 5) are consistent with the reduced carbon assimilation associated with the stomatal closure described above, reinforcing the case for RDI to sustain carbon assimilation through to maturity. Beyond terpene preservation, preventing prolonged stomatal closure through RDI would also reduce the risk of whole-canopy carbon starvation—the mechanism by which near-isohydric varieties eventually develop diffuse leaf yellowing under extended drought [87]—thereby protecting canopy integrity and the vine’s capacity for subsequent seasons.

4.4.2. Fetească Neagră

Fetească neagră requires targeted interventions to reduce thermal stress in the cluster zone. Physical approaches such as shading nets, kaolin or zeolite applications can reduce cluster temperatures [88,89], partially mitigating heat-induced inhibition of anthocyanin biosynthesis and tannin maturation; high temperatures are known to suppress expression of key anthocyanin structural genes and MYB-family regulators in grape skin [10,94]. Since ABA is the primary hormonal signal for stomatal closure [95,96], foliar ABA applications could plausibly shift anisohydric behaviour toward a more water-conserving pattern; this effect has been specifically characterised in near-isohydric (Montepulciano) versus anisohydric (Sangiovese) cultivars, where foliar ABA accumulation and stomatal closure dynamics differ markedly between the two behaviours [97]. Biochemical elicitors such as methyl jasmonate and benzothiadiazole (BTH) have also been shown to enhance antioxidant enzyme activity and modulate protective metabolic responses under water deficit in grapevine, offering a complementary chemical strategy for priming stress tolerance in Fetească neagră [98,99]. Foliar or root-applied ethanol has recently been shown to reduce transpiration and slow the depletion of plant-available soil water in Vitis vinifera cuttings under progressive drought, without increasing leaf senescence [100]. Practical application would require optimisation of dose (lower concentrations for irrigation-based delivery than for foliar spraying) and timing, but the low cost and existing on-site availability of this compound make it an attractive candidate for managing Fetească neagră’s anisohydric water loss. Acetic acid has similarly been shown to activate ABA- and jasmonic acid-mediated drought avoidance responses, though this has so far only been demonstrated in other crop species [101].

Coupling these interventions with RDI during the ripening period could plausibly help regulate sugar accumulation and reduce berry shrivelling risk. The high polyphenolic and anthocyanin potential recorded in Fetească neagră at harvest (72.9 AU Kg⁻¹; 270.9 mg L⁻¹ anthocyanins, total potential 811.3 mg Kg⁻¹, Table 5) could represent a quality asset worth protecting, pending validation of these approaches, before it is lost to the irreversible shrivelling pathway described, which can be triggered at véraison.

4.5. Limitations

Direct measurement of leaf and stem water potential was beyond the scope of this study; the near-isohydric and anisohydric positioning discussed throughout is therefore inferred from gas exchange parameters (A, g_s, E, C_i, WUE), assimilatory pigment indices (Chl a, Chl b, C+X, Chl/(C+X), Chl a/b), and leaf dry matter content. This approach is well-established in the grapevine physiology literature [25,26], but remains an indirect proxy for the hydraulic regulation that formally defines this continuum. As a result, both cultivars are placed toward, rather than at, its extremes. This conclusion is supported empirically by observations that even canonical benchmark varieties, such as Grenache, Syrah, Chardonnay and Malbec, do not consistently maintain their classified positions across different climatic and edaphic conditions [29]. It should also be acknowledged that measurements were collected during a single, climatically exceptional season with three biological replicates per cultivar; the physiological values reported here are best read as a detailed profile of how these cultivars responded to the 2024 semi-arid conditions, rather than as stable trait parameters. Multi-year monitoring across seasons of varying severity is planned as follow-up work to establish whether the water use strategies described here are stable cultivar traits or responses specific to the exceptional drought conditions of 2024. Direct pre-dawn and midday leaf and stem water potential measurements are likewise planned in this follow-up work to directly test the hydraulic mechanisms proposed here, which in the present study remain inferences grounded in gas exchange proxies and the established grapevine physiology literature rather than direct measurements. The berry weight and sugar content dynamics approach introduced in this study (Section 2.3.6) will be extended with replicated sampling in future seasons to characterise ripening trends with greater confidence. We also note that leaf gas exchange was measured only at flowering, berry growth, and véraison, and not during the late-ripening period in which the sugars expressed as g berry⁻¹ (Supplementary Figure S2) suggest that passive sugar concentration becomes dominant; consequently, whether photosynthetic activity remained elevated specifically during this later window, as assumed in Section 4.2, was not directly tested and represents a further target for the planned follow-up study.

5. Conclusions

This study characterised the contrasting physiological drought responses of two Romanian autochthonous grapevine cultivars under semi-arid field conditions (DMI: 23.6; HC: 0.75) during the exceptionally hot and dry 2024 growing season. PCA and LMM analyses consistently resolved Șarba’s physiological profile as consistent with a near-isohydric, water-conserving response, characterised by early stomatal closure, high WUE, and maintained chlorophyll at véraison, while Fetească neagră’s profile was consistent with anisohydric behaviour, sustaining high gas exchange rates and carotenoid-based photoprotection throughout the season.

Premature turgidity loss and senescence of basal leaves, visually observed in Fetească neagră vineyards across Romania in recent years, are, to the best of our knowledge, formally documented here for the first time, and highlight the physiological vulnerability of anisohydric strategies under prolonged severe drought.

These findings provide a physiological basis for cultivar-specific drought management strategies, supporting the use of regulated deficit irrigation for Șarba and combined elicitor treatments with supplemental irrigation for Fetească neagră. Under the climate change trajectory projected for Romanian viticultural regions, the isohydric–anisohydric continuum should be integrated into varietal selection and irrigation management frameworks to safeguard both yield and wine quality.

Extending these observations across multiple seasons and drought intensities, with additional molecular-level investigation, would help clarify the basis of the observed cultivar divergence.