The Role of Water and Atmospheric CO₂ on the δ¹³C Value of Sugars of Grape Must

Abstract

Climatic parameters influence the δ¹³C value of sugar in grape must. With the aim of investigating this dependence, grape must samples were collected from two viticultural Italian areas (Oltrepò Pavese, Lombardia region and Illasi–Mezzane, Veneto region), which share similar soil mineralogical compositions. Water uptake by the plant is the primary factor affecting the δ¹³C values of sugar: the greater the water availability, the lower the δ¹³C value. This is supported by a correlation between the δ¹³C values and the climatic water balance (BIC_c), which is defined as the difference between daily rainfall and crop evapotranspiration. Pre-harvest atmosphere was also sampled at both sites to determine its concentration and δ¹³C value. Using the Farquhar model, enrichment factors and ε_CO2-sugar were calculated. A moderate correlation was found between cumulative rainfall and the associated values of the enrichment factor: approximately 60% of the variation in sugar δ¹³C can be attributed to water uptake and to the δ¹³C values of atmospheric CO₂. Rainfall alone showed an even stronger correlation with ε_CO2-sugar, suggesting that water availability is the dominant factor influencing the sugar δ¹³C.

1. Introduction

Plants have different metabolic mechanisms: C3, C4, and CAM. C3 plants represent 85% of all higher plants; they convert CO₂ directly into 3-phosphoglyceric acid (PGA), an organic compound with three carbon atoms. In C4 and CAM plants, CO₂ is turned into oxaloacetate (an intermediate of four carbon atoms) before transforming into PGA. The differences between C4 and CAM plants are the following: CAM plants store CO₂ during the night, and oxaloacetate is turned into malic acid before conversion into PGA [1]. These metabolic processes lead to different carbon isotope composition in the formation of organic compounds.

The parameter $\mathsf{\delta }$ is used to express the isotope composition referred to in international reference material. For carbon we have

$$\mathsf{\delta }{}^{13}\mathrm{C}={\displaystyle \frac{{\left({}^{13}\mathrm{C}/{}^{12}\mathrm{C}\right)}_{\mathrm{s}}}{{\left({}^{13}\mathrm{C}/{}^{12}\mathrm{C}\right)}_{\mathrm{V}\mathrm{P}\mathrm{D}\mathrm{B}}}}-1$$

(1)

where ${\left({}^{13}\mathrm{C}/{}^{12}\mathrm{C}\right)}_{\mathrm{s}}$and ${\left({}^{13}\mathrm{C}/{}^{12}\mathrm{C}\right)}_{\mathrm{V}\mathrm{P}\mathrm{D}\mathrm{B}}$ are the ratios between the less abundant isotope, ${}^{13}C$, and the most abundant isotope, ${}^{12}\mathrm{C},$ present in the substance s of interest and in the primary reference material VPDB [2]. The range of the values of this parameter is different for each plant group: C3 plants vary from −21‰ to −35‰ [3], C4 plants from −9‰ to −16‰ [4], and CAM plants from −15‰ to −20‰ [5].

Vine is a C3 plant with several stages of the annual growth cycle: shoot and inflorescence development, flowering, berry development, ripening, and senescence [6]. During berry development (50 to 120 days before the harvest), the sugar starts to translocate into the new young berries. This process depends on climate conditions and the latitude of the vineyard. According to the sigmoidal curve described by Coombe [7], the maximum rate of carbohydrate translocation occurs between 50 and 70 days before the harvest, as the acids convert into sugars.

During the photosynthesis reaction, atmospheric CO₂ is captured by the stomata of leaves and transformed into carbohydrates. It occurs in different steps: the “light” and the “dark” reaction [8]. In the “light” reaction, water is split using solar radiation energy with production of oxygen (Equation (2)), protons, and electrons. In the “dark” reaction, protons and electrons are used to transform CO₂ into carbohydrates (Equation (3)) [9]:

$${12\mathrm{H}}_2{\mathrm{O}}_{\left(\mathrm{l}\right)}\to {24\mathrm{H}}^{+}+{24\mathrm{e}}^{-}+{6\mathrm{O}}_{2\left(\mathrm{g}\right)}$$

(2)

$$6\mathrm{C}{\mathrm{O}}_{2\left(\mathrm{g}\right)}+24{\mathrm{H}}^{+}+24{\mathrm{e}}^{-}\to {\mathrm{C}}_6{\mathrm{H}}_{12}{\mathrm{O}}_{6\left(\mathrm{a}\mathrm{q}\right)}+6{\mathrm{H}}_2{\mathrm{O}}_{\left(\mathrm{l}\right)}$$

(3)

The overall reaction is

$${6\mathrm{C}{\mathrm{O}}_{2\left(\mathrm{g}\right)}+6\mathrm{H}}_2{\mathrm{O}}_{\left(l\right)}\to {\mathrm{C}}_6{\mathrm{H}}_{12}{\mathrm{O}}_{6\left(\mathrm{a}\mathrm{q}\right)}+{6\mathrm{O}}_{2\left(\mathrm{g}\right)}$$

(4)

The standard Gibbs free energy of the overall reaction (Equation (4)) is ($∆{\mathrm{G}}_{\mathrm{r}}^0$) $\approx$ + 2870 kJ/mol (P = 1 bar, T = 298.15 K). The reaction is not spontaneous; it needs light energy to proceed [10,11]. Water, light, and atmospheric carbon dioxide are fundamental ingredients for photosynthesis. Water, light, and atmospheric carbon dioxide levels can influence sugar production in grape berries by affecting photosynthesis and metabolic enzyme activity. Different light wavelengths regulate water uptake, element absorption, and sugar metabolism. Limited light (especially yellow, blue, and red) reduces sugar production (decrease in enzyme activity) [12]. High CO₂ can enhance the photosynthetic rate and may also lead to a decline in photosynthetic efficiency due to reduced nitrogen availability [13]. Water availability also plays a major role, with higher levels promoting greater sugar accumulation and causing more negative δ¹³C values. The last topic will be discussed in the next paragraph.

The total amount of sugar in leaves (before they are transported to the grape berries) depends on various environmental factors like vine water status, local time, hourly radiation, and the period of growth considered [14,15,16]. Furthermore, the δ¹³C value of sugar in grape berries (and consequently in grape musts) depends on climate parameters. Taskos et al. [17] found that the carbon isotope composition of grapevine is affected by water and nitrogen supply. Additionally, Brillante et al. [18] found a good relationship between the δ¹³C of grape must with stomatal conductance (g_s), intrinsic water use efficiency (WUEi), and stem water potential (ψ_stem). A significant relationship has also been observed between the soil water content (SWC) and the δ¹³C value of grape must [19,20]. According to Kolesnov et al. [19], a δ¹³C value below −26‰ is indicative of the absence of water stress, whereas values exceeding −21.5‰ indicate severe water stress. Gómez-Alonso et al. [21] demonstrated significant differences in sugar δ¹³C values between irrigated and non-irrigated vineyards; non-irrigated vineyards show higher sugar δ¹³C values in grape must in comparison with the irrigated ones. This highlights the importance of water for plants. Other studies also confirm the relationship between rainfall and water adsorbed by vineyards [22,23,24]. Furthermore, Zhang et al. [25] stated that δ¹³C values depend not only on climate parameters, the isotopic composition of atmospheric CO₂ and the type of vegetation but also on atmospheric CO₂ concentration.

A relation between the atmospheric pressure ($p_a$), the intracellular pressure $\left(p_i\right),$ and the enrichment factor, Δ, has been proposed [26]:

$$∆ =a+(b-a)·{\displaystyle \frac{p_i}{p_a}}$$

(5)

where $a$ and $b$ are enrichment fractionation factors; the former due to the diffusion of CO₂ in air through the stomatal pore, and the latter due to the reaction of carboxylation by Rubisco. For the sugar produced by leaves of C₃ plants, the value of $a$ is 4.1‰, whereas the value of $b$ ranges from 24 to 25.5‰ [26,27]. It is noteworthy that $∆$is an inappropriate symbol of the “isotope enrichment factor”, which is indicated as ε since the middle of the last century in chemistry and geochemistry [2]:

$$\mathsf{\epsilon }={\displaystyle \frac{{\mathsf{\delta }}_{\mathrm{a}}-{\mathsf{\delta }}_{\mathrm{p}}}{1+{\mathsf{\delta }}_{\mathrm{p}}}}$$

(6)

where ${\mathsf{\delta }}_{\mathrm{a}}$ refers to the carbon of the atmospheric CO₂ and ${\mathsf{\delta }}_{\mathrm{p}}$ to the plant carbon. In our case, ${\mathsf{\delta }}_{\mathrm{p}}$ corresponds to the carbon isotopic composition of the sugar of the grape must.

For all these reasons, it is important to study the annual variation in the carbon isotope value of grape-must sugar. Since these values are linked to intrinsic water use efficiency (iWUE) in plants and other climate-related parameters, they can help us better understand plant response strategies to climate change [28,29].

The grape must contain water, sugar, organic acids, nitrogen compounds, polysaccharides, minerals, polyphenols, vitamins, and aromatic compounds. Sugar is the second most abundant compound in grape juice after water. The maximum values of the interval of each compound in the must are as follows [30]: 75% wt of water, 22% wt of sugar, and 1.5% wt of organic acids.

The aims of this paper are the following:

(i)

Identification of meteorological/climate parameters, which primarily influence the isotope composition of sugar in two delimited areas of the northern Italy;

(ii)

Definition of a linear correlation between meteorological/climate parameters and δ¹³C values of the must;

(iii)

Sampling of air in the two different areas to estimate the carbon isotope composition and concentration of CO₂ in the sampled atmosphere, and the evaluation of the enrichment factor (ε) between the carbon isotope composition of the atmospheric CO₂ and that of the grape-must sugar (Equation (6)).

2. Materials and Methods

Since the occurrence of phyllosilicate with an expandable lattice (smectite) may play an important role in retaining water during dry periods [31], the mineralogy of the soil has been investigated by X-ray powder diffraction.

2.1. Areas of Investigation

Two distinct Italian areas were selected for this study:

(1)

Oltrepò Pavese, located in the south of the Lombardia Region (Pavia province); in this area, six different sites were considered: Santa Maria della Versa, Cigognola, Canevino, Montalto Pavese, Montebello della Battaglia, and Borgoratto Mormorolo. Mineral investigation of the soil material revealed the presence of calcite, quartz, and feldspar. The clays minerals identified were smectite, illite, chlorite, and kaolinite.

(2)

The Illasi and Mezzane area, which is in the northern part of the province of Verona. Mineralogical investigation of the soil reveals the presence of calcite, dolomite, quartz, feldspars, kaolinite, chlorite, illite, and smectite.

2.2. Sampling

The Oltrepò Pavese grape varieties investigated are the following: Rhenish Riesling, Croatina, Pinot Noir, Barbera, Italic Riesling, and Cabernet. The Illasi–Mezzane grape varieties are as follows: Rhenish Riesling, Manzoni grape, Garganega, Croatina, Corvivone, Rondinella, and Syrah.

2.2.1. Oltrepò Pavese

During the 2021 and 2022 harvests, a bunch of grapes were collected weekly (Figure 1) over a four-week period. Samples were taken from one vineyard at Borgoratto Mormorolo, Canevino, Cigognola, and Santa Maria della Versa; from two vineyards at Montebello della Battaglia; and from three vineyards in Montalto Pavese. It is noteworthy that (i) for both years 2021 and 2022, vineyards were not irrigated manually by farmers and then, the plants adsorbed only the meteoric water that fell before the harvest; (ii) the bunch of grapes were collected from the same geographical points each year. A meteorological station is present in each locality that monitors air temperature, solar radiation, relative humidity, rainfall amount, and wind speed.

During the three months prior to the 2022 harvest (June, July, and August) air samples were collected from the vineyard area to determine δ¹³C and the concentration (ppm) of the atmospheric CO₂. Sampling was performed once per month in the sites of Santa Maria della Versa and Cigognola.

2.2.2. Illasi–Mezzane

For this area, vineyards are dislocated between Illasi and Mezzane. Unlike the Oltrepò Pavese area, samples were collected on different dates and not systematically. Bunches of grapes were collected from nine non-irrigated vineyards in 2021 and from eight manually irrigated vineyards in 2022 (Figure 1). The meteorological parameters were monitored by a climatic station in Illasi (ARPAV, Agenzia regionale per la Prevenzione e Protezione Ambientale del Veneto). Before the 2022 harvest, at the beginning of June, July, and August, air was sampled in the vineyard area to determine δ¹³C and the concentration of the atmospheric CO₂.

2.3. Analytical Methods

2.3.1. Carbon Isotope Composition of Atmospheric CO₂

The concentrations (ppm) of CO₂ in air and relative δ¹³C values were measured using a WS-CRDS (Wavelength-Scanned Cavity Ring Down Spectroscopy) with Picarro G2201-i analyzer (Corporate, Picarro, Inc., Santa Clara, CA, USA) (pumping rate: 25 mL min⁻¹). Air samples were collected using 10L Tedlar^® gas sampling bags filled directly in the field using a syringe and three-way valve. Calibration was performed in the lab using Picarro (Corporate, Picarro, Inc., Santa Clara, CA, USA). At the beginning of every measuring period, calibration was performed using the following standards produced, on request, by Air Liquide: (i) 380, 500, and 1000 ppm vol CO₂; (ii) −44, −5, and 2‰ vs. VPDB δ¹³C-CO₂. These internal standards were originally inter-calibrated by using the IAEA international standards and cross-analyzed with other isotopic laboratories (National Institute of Geophysics and Volcanology (INGV), Naples Headquarters, and the Institute of Geosciences, Goethe University Frankfurt am Main, Germany). The standard used were recommended by the National Institute of Health for CO₂ monitoring [32]. The analytical errors for GC and IRMS analysis of standards were 5% and ±0.06‰, respectively. The Picarro G2201-i Analyzer is an instrument designed for continuous acquisition; it can perform approximately one measurement per second. The samples of this work were collected in specific bags for sampling the gas phases and analyzed for a minimum of 5 min. This procedure allows us to obtain at least 300 measurements. The precision was within 0.2 ppm for CO₂ and 0.16‰ for δ¹³C-CO₂.

2.3.2. Dry Must

Each sample of the collected bunches of grapes was pressed to obtain grape must, filtered using a 0.45 µm filter, and immediately frozen to prevent fermentation. Subsequently, the samples were placed in a freeze dryer for three days to remove water and obtain solid residue. As mentioned earlier, sugar is the second most abundant compound in grape juice after water [30]; thus, the solid obtained after freeze-drying primarily consists of sugar.

The analyses were performed using the IRMS instrument Delta V Advantage (Thermo Scientific, Bremen, Germany), hyphenated with the Thermo Fisher Flash HT plus Elementar Analyzerms in Dynamic Flash Combustion (DFC) mode for ¹³C/¹²C measurements (Thermo Scientific, Bremen, Germany). Gas separation was achieved with the IRMS Separation column (NC separation column, Thermo Scientific, Bremen, Germany), operated at 45 °C for δ¹³C analysis. The EA-device, equipped with the autosampler MAS 200R for solids (Thermo Scientific, Bremen, Germany), was placed in-line with the MS in a high purity He flow (purity 6.0, Rivoira Gas) via the ConFlo IV (Thermo Scientific, Bremen, Germany), which is capable of alternately addressing both sample and reference gases. The carrier flow and reference flow were both at 100 $\mathrm{m}\mathrm{L}·{\mathrm{m}\mathrm{i}\mathrm{n}}^{-1}$. The reactor temperature was 1020 °C. The calibration was performed using the following reference materials: NBS22 (natural oil, δ¹³C = (−30.03 ± 0.05)‰), IAEA-CH-6 (sucrose, δ¹³C = (−10.45 ± 0.04)‰), and USGS24 (graphite, δ¹³C = (−16.05 ± 0.04)‰). Additionally, to verify the calibration quality, IAEA-CH-7 (polyethylene, δ¹³C = (−32.15 ± 0.05)‰) was used as the control standard, recalculating its value. All reference materials used come from International Atomic Energy Agency (IAEA), Vienna, Austria. Prediction uncertainty was inferred (probability 95%) according to statistical methods described by [33,34]. At a 0.05 level of significance, the prediction uncertainty value is ±0.26‰.

3. Results and Discussion

3.1. Best Relationship Between δ¹³C Sugar Values and Climate Parameters

Numerous studies have demonstrated the influence of climate parameters on the δ¹³C values of the plant organic compound. This study investigated the main meteorological/climate parameter influencing the δ¹³C values of the must sugar. We determined the best linear correlation between each meteorological parameter (air temperature, relative humidity, solar radiation, rainfall, actual vapor pressure eₐ, and wind speed) and the δ¹³C values. Instead of using parameter values from the harvest date, we averaged values from several days before harvest. The optimal averaging period was determined by maximizing the R² value and minimizing the standard error of the linear regression between δ¹³C and the averaged parameter values. A custom Matlab script performed this iterative calculation until a predefined maximum number of days was reached (Matlab, 2020; version 2020a; Natick, Massachusetts, USA: The MathWorks Inc.). In this case, the δ¹³C values obtained and used in this research were associated with the total sugar since they were the most heaping organic compound in the grape musts. For this reason, attempts have been made to verify the function of the possible concentration (g/L) [35], the possible carbon isotope value [36], and how tartaric acid and malic acid could influence on the δ¹³C sugar value. It has been found that the variations are included in the uncertainty. For this reason, it is possible to say that the fingerprint of the organic compound of grapes must mostly depend on fructose and glucose.

3.1.1. Harvest 2021

The Oltrepò samples exhibit a good linear correlation between the mean of the actual vapor pressure (Table S1), the sugar δ¹³C (Table S3), and ${\mathrm{e}}_{\mathrm{a}}$ of the atmospheric water [37] for 101 days before the harvest (Figure 2).

The actual vapor pressure, ${\mathrm{e}}_{\mathrm{a}}$, is calculated using the following equation:

$${\mathrm{e}}_{\mathrm{a}}={\displaystyle \frac{{\mathrm{e}}^0\left({\mathrm{T}}_{\mathrm{m}\mathrm{i}\mathrm{n}}\right)·\left(\frac{{\mathrm{R}\mathrm{H}}_{\mathrm{m}\mathrm{a}\mathrm{x}}}{100}\right)+{\mathrm{e}}^0\left({\mathrm{T}}_{\mathrm{m}\mathrm{a}\mathrm{x}}\right)·\left(\frac{{\mathrm{R}\mathrm{H}}_{\mathrm{m}\mathrm{i}\mathrm{n}}}{100}\right)}{2}}$$

(7)

where RH_max and RH_min are the maximum and the minimum daily value of relative humidity, respectively; ${\mathrm{e}}^0\left({\mathrm{T}}_{\mathrm{m}\mathrm{i}\mathrm{n}}\right)$ and ${\mathrm{e}}^0\left({\mathrm{T}}_{\mathrm{m}\mathrm{a}\mathrm{x}}\right)$ are the saturation vapor pressure related to air minimum and maximum temperature calculated by the following equation:

$${\mathrm{e}}^0\left(\mathrm{T}\right)=0.6108\mathrm{e}\mathrm{x}\mathrm{p}\left({\displaystyle \frac{17.27·\mathrm{T}}{\mathrm{T}+237.7}}\right)$$

(8)

δ¹³C and ${\mathrm{e}}_{\mathrm{a}}$ exhibit an inverse relationship (Figure 2).

Actual vapor pressure is linked to vapor pressure deficit (VPD), which is defined as the difference between the mean saturation vapor pressure $\left({\mathrm{e}}_{\mathrm{s}}\right)$ and the actual vapor pressure, ${\mathrm{e}}_{\mathrm{a}}$. VPD plays a crucial role in regulating the opening and closure of leaf stomata. Specifically, high VPD values indicate low relative humidity and, thus, an increase in stomatal closure. Conversely, low VPD values are associated with high relative humidity and reduced stomatal closure [38,39]. Stomata controls the inflow of CO₂ and the release of water into the atmosphere. Solar radiation, rainfall amount, and wind speed values were also used to check for a linear correlation with the δ¹³C values, but no statistically significant relations were found (p-value > 0.05).

Regarding Illasi–Mezzane, no relationship was found between the climatic parameters (air temperature, relative humidity, ${\mathrm{e}}_{\mathrm{a}}$, solar radiation, wind speed, rainfall amount) and sugar δ¹³C values. This is probably due to the small sample size.

3.1.2. Harvest 2022

Oltrepò Pavese

The Oltrepò Pavese samples show an inverse linear trend between the cumulative rainfall over the 109 days preceding harvest (Table S6) and sugar δ¹³C values (Figure 3, black points). The δ¹³C values are divided into two groups based on the amount of rainfall. Samples from Cigognola and Santa Maria della Versa, which experienced exceptionally high rainfall amounts during 2022, form the cluster in the lower right of Figure 3. These samples have a significantly lower overall δ¹³C median compared with the others (p-value < 0.001 Mann–Whitney test). The statistical investigation of the carbon isotopic values of sugar from different grape varieties did not reveal any significant differences (p_{same average} = 0.42, Kruskal–Wallis test).

The same relationship has been observed for the Illasi–Mezzane samples of 2022 (Figure 4), albeit with a modest linear correlation found for the 99 days preceding harvest in this area (Tables S8 and S9). This time interval includes the onset of sugar translocation into young berries [6,7].

This suggests that the amount of rainfall may be linked to the variability of δ¹³C. In fact, no correlation was found with the actual vapor pressure (ea), similarly to the 2021 harvest. Most probably, it depends on the consequently crucial role water plays in sugar production during photosynthesis. In fact, water fulfils multiple functions throughout the photosynthetic process. In the first step (light-dependent phase), water undergoes photolysis-splitting driven by solar radiation-providing electrons that are ultimately used to reduce carbon oxide:

$${\mathrm{H}}_2\mathrm{O}\to 2{\mathrm{H}}^{+}+2{\mathrm{e}}^{-}+{\displaystyle \frac{1}{2}}{\mathrm{O}}_2$$

(9)

The second step (the dark- or light-independent phase) does not depend on solar radiation. Instead, the driven force is furnished by the hydrolysis energy of ATP (adenosine triphosphate) to ADP (adenosine diphosphate) and by the oxidation of NADPH (nicotinamide adenine dinucleotide phosphate hydrogen) to NADP⁺ (nicotinamide adenine dinucleotide phosphate). The latter, provides high-energy electrons required for the conversion of carbon dioxide into glyceraldehyde-3-phosphate (G3P), a three-carbon sugar precursor to glucose [40].

Water plays an essential role in these steps. Its functions are summarized below:

1

NAPDH formation occurs using electrons and proton released by the splitting of water during the light-dependent phase, as shown in the following reaction [41]:

$$\mathrm{N}\mathrm{A}\mathrm{D}{\mathrm{P}}^{+}+2{\mathrm{H}}^{+}+2{\mathrm{e}}^{-}\to \mathrm{N}\mathrm{A}\mathrm{D}\mathrm{P}\mathrm{H}+{\mathrm{H}}^{+}$$

(10)

2

Energy released by ATP hydrolysis involves water, as shown in the following reaction [42]:

$$\mathrm{A}\mathrm{T}\mathrm{P}+{\mathrm{H}}_2{\mathrm{O}}_{\left(\mathrm{l}\right)}⟶\mathrm{A}\mathrm{D}\mathrm{P}+{\mathrm{P}}_{\mathrm{i}}+{\mathrm{H}}^{+}$$

(11)

3

Finally, water is a medium for solute transport throughout the plant.

It is clear that water is the limiting reagent in the photosynthetic reaction, especially considering that atmospheric CO₂ concentration is increasing due to anthropogenic input (latest detection at Mauna-Loa 426.4 ppm vol, data access 27 June 2025, https://gml.noaa.gov/ccgg/trends/global.html).

During plant development, a significant portion of the total available water (soil and absorbed water) is lost through evapotranspiration. Taking this into account, we investigated whether a correlation exists between crop evapotranspiration, ETc, and sugar δ¹³C values. To assess this, we used a climatic indicator known as hydroclimatic balance (BIC), which reflects the balance between the rainfall (mm) and potential evapotranspiration (ET₀):

$$\mathrm{B}\mathrm{I}\mathrm{C}=(\mathrm{m}\mathrm{m} \mathrm{o}\mathrm{f} \mathrm{r}\mathrm{a}\mathrm{i}\mathrm{n}-{\mathrm{E}\mathrm{T}}_0)$$

(12)

To relate the water availability more directly to plant water needs, ET₀ was replaced with crop evaporation (ET_c):

$$\mathrm{B}\mathrm{I}{\mathrm{C}}_{\mathrm{c}}=(\mathrm{m}\mathrm{m} \mathrm{o}\mathrm{f} \mathrm{r}\mathrm{a}\mathrm{i}\mathrm{n}-{\mathrm{E}\mathrm{T}}_{\mathrm{c}})$$

(13)

Both BIC_c and rainfall were calculated daily. ET₀ and ET_c values were estimated according to FAO guidelines the [37,43]. ET₀ was calculated according to the following equation (Equation (14)):

$$\mathrm{E}{\mathrm{T}}_{\mathrm{O}}={\displaystyle \frac{0.408·∆·({\mathrm{R}}_n-\mathrm{G})+\mathsf{\gamma }·900 \mathrm{T}+273·{\mathrm{u}}_2·({\mathrm{e}}_s-{\mathrm{e}}_a)}{∆+\mathsf{\gamma }·(1+0.34·{\mathrm{u}}_2)}}$$

(14)

where R_n is the net radiation at the crop surface [MJ m⁻² day⁻¹], G is the soil heat flux density [MJ m⁻² day⁻¹], e_a is the actual water vapour pressure [kPa], e_s is the saturation vapour pressure [kPa], T is the mean daily air temperature [°C] at 2 m from the soil, Δ is the slope vapor pressure curve [kPa °C⁻¹], γ is the psychrometric constant [kPa °C⁻¹], and u₂ is the wind speed [m s⁻¹] at 2 m from the soil.

Daily crop evapotranspiration was calculated using equation reported below:

$$\mathrm{E}{\mathrm{T}}_{\mathrm{C}}={\mathrm{K}}_{\mathrm{C}}\mathrm{E}{\mathrm{T}}_0$$

(15)

where K_C is the crop coefficient and ET₀ is the daily potential evapotranspiration. K_c reflects the effect of the characteristics that distinguish a typical field crop from the grass of reference. Consequently, different crops have different K_C coefficients. Furthermore, the value of K_C depends on the following: (i) four growth stages: initial, crop development, mid-season, and late season; (ii) the stage length (in days), which depends on the latitude; (iii) the type of crop; and (iv) the height of the crop (h_P).

In this study, the crop is grape vines. The values of K_{C, initial-season}, K_{C, middle season}, and K_{C, end-season} are tabulated (0.30, 0.70, and 0.45, respectively) [37,43]. The meteorological conditions for which the tabulated values were obtained are the following: minimum daily relative humidity (h_min) = 45% and wind speed = 2 m/s (u₂) (measured at 2 m above the ground). If h_min and/or u₂ is/are different from the values reported above, the K_C values must be adjusted using the following equation [37,43]:

$${\mathrm{K}}_{\mathrm{C}\left(\mathrm{a}\mathrm{d}\mathrm{j}\mathrm{u}\mathrm{s}\mathrm{t}\mathrm{e}\mathrm{d}\right)}={\mathrm{K}}_{\mathrm{C}\left(\mathrm{t}\mathrm{a}\mathrm{b}\mathrm{u}\mathrm{l}\mathrm{a}\mathrm{t}\mathrm{e}\mathrm{d}\right)}+[0.04\left({\mathrm{u}}_2-2\right)-0.004\left({\mathrm{h}}_{\mathrm{m}\mathrm{i}\mathrm{n}}-45\right)]·{\left({\displaystyle \frac{{\mathrm{h}}_{\mathrm{p}}}{3}}\right)}^{0.3}$$

(16)

The period of interest is between the mid-season and late season, when maturity begins [37,43]. The late season may reach 80 days [37,43]. Unlike K_{C(initial-season)}, K_{C(middle-season),} and K_{C(end-season)} values, the K_{c(late-season)} value is the result of an interpolation between the K_{C(middle-season)} and K_{C(end-season)}. To estimate K_{c(late-season)}, we used Equation (16), starting from the tabulated K_{C(end-season)} value (0.45) and replacing the values of h_min and u₂ with experimental daily data. If the resulting value was equal to or greater than 0.45, it was retained; otherwise, the original value of 0.45 was maintained. In this way, we aimed to simulate values ranging between K_{C(middle season)} (0.70), and K_{C (end-season)} (0.45).

For the Oltrepò Pavese, a linear correlation was observed between the average BICc, calculated over the 61 days preceding each harvest date (Table S7), and the corresponding δ¹³C of the must (Figure 5).

Figure 5 illustrates the crucial role of balance between the water amount (mm of rain) and crop evapotranspiration (ET_c) in influencing the isotope composition of grape sugar. When ETc exceeds rainfall (negative BICc), sugar becomes enriched in ¹³C. Conversely, when rainfall exceeds ETc, the δ¹³C value decreases. This trend reinforces importance of water availability in the photosynthetic process. Furthermore, the regression equation shown in Figure 5 indicates that, when the independent variable BICc equals zero (that is when rainfall matches ETc), the δ¹³C value of sugar is

$${\mathsf{\delta }}^{13}{\mathrm{C}}_{\mathrm{V}\mathrm{P}\mathrm{D}\mathrm{B}}=(-26.96\pm 0.75)\mathrm{‰}$$

(17)

The error term was calculated as the standard error formula of intercept at 95% confidence level. As shown in Figure 3 and Figure 5, the coefficient of determination, R², is greater in the relationship between BICc and δ¹³C (R² = 0.50) than for the relationship between cumulative rainfall alone and δ¹³C (R² =0.45). This indicates that (i) water loss though evapotranspiration significantly influences the isotopic composition of sugar and (ii) for R² value of 0.5, 50% of the variance of δ¹³C can be explained by the actual water availability.

The remaining 50% of the variability may be related to physiological or environmental factors. For example, higher leaf nitrogen concentrations have been associated with an increase in δ¹³C values [18]. Additionally, an increase in stromal Mg²⁺ concentration and pH have been shown to promote the ATP synthesis [44,45], which provides energy during the dark phase of the photosynthesis. The isotopic composition of atmospheric CO₂ also contributes to this variability.

The same calculation was performed for the Illasi–Mezzane samples (Figure 6), using the mean δ¹³C value in cases where multiple samples were collected on the same harvest date (Tables S8 and S9).

In Figure 6, the observed trend is similar to that found for the Oltrepò. In both cases, the period preceding harvest that shows the strongest linear correlation is nearly identical: 61 days in Oltrepò and 60 days at Illasi–Mezzane (Table S7). For Illasi–Mezzane, when BICc is equal to zero, the sugar δ¹³C value is

$${\mathsf{\delta }}^{13}{\mathrm{C}}_{\mathrm{V}\mathrm{P}\mathrm{D}\mathrm{B}}=\left(-24.90\pm 0.46\right)\mathrm{‰}$$

(18)

This value is slightly higher than in the Oltrepò Pavese. Likely, this is due to the smaller sample size. The error represents the standard error of intercept at a 95% confidence level.

Lawlor and Cornic [46] demonstrated that a water deficit in plants inhibits ATP synthesis, which consequently reduces ribulose-1,5-bisphosphate (RuBP) synthesis, the principal CO₂ acceptor in the Calvin cycle. This reduction limits the potential CO₂ assimilation rate and contributes to variations in the isotopic composition of sugar. Moreover, Gilbert et al. [47] highlighted that one important factor influencing the intramolecular carbon isotope distribution in sucrose is the kinetic isotope effect associated with decarboxylation or other steps that generate respiratory metabolites.

In the 2022 harvest, no statistically significant relationship was found between δ¹³C values and the actual vapor pressure. Unlike in 2021, rainfall in 2022 was more abundant, particularly at the Santa Maria della Versa and Cigognola sites. Gross primary productivity (GPP), defined as the total amount of carbon compounds produced by plant photosynthesis within an ecosystem, is influenced by both soil water content (SWC) and vapor pressure deficit (VPD). Regarding the relative roles of SWC and VPD, in the literature it has been found that VPD is the dominant factor driving drought stress on ecosystem productivity, while SWC becomes more relevant under dry soil conditions [48]. For this reason, we suggest that the lower rainfall in 2021 led to drought stress conditions in the plants, making actual vapor pressure (closely related to VPD) the main factor influencing the δ¹³C values. In contrast, in 2022, the higher amount of rainfall reduced drought stress (Figure 3), diminishing the impact of VPD. Consequently, water availability likely played a more important role than actual vapor pressure (or VPD) in determining the δ¹³C values during the 2022 harvest.

3.1.3. Relationship Between δ¹³C in Sugar and in Atmospheric CO₂

During the three months preceding the 2022 harvest (June, July, and August), air samples were collected from vineyard sites to determine the δ¹³C values and the concentration (ppm vol) of the atmospheric CO₂. Sampling was conducted once per month at 12:00 a.m., when net photosynthesis reached its maximum [49], at the following locations: Santa Maria della Versa, Cigognola for the Oltrepò Pavese area, Illasi (three samples), and Mezzane (three samples as the other sites, although one was lost due to bag failure).

Table 1. Isotopic composition of atmospheric CO₂ and its concentration (in ppm) for Cigognola, Santa Maria della Versa, Illlasi and Mezzane during June, July, and August 2021 and 2022.

	Oltrepò Pavese Area				Illasi–Mezzane Area
	Santa Maria della versa		Cigognola		Illasi		Mezzane
	δ13CVPDB	ppm vol	δ13CVPDB	ppm vol	δ13CVPDB	ppm vol	δ13CVPDB	ppm vol
June	−10.45 (±0.69)	468.55 (±0.95)	−8.79 (±067)	431.94 (±3.47)	−11.7 (±0.14)	428.00 (±0.14)	−10.9 (±0.14)	412.00 (±0.14)
July	−9.21 (±0.77)	425.41 (±0.14)	−9.01 (±0.88)	431.68 (±1.01)	−12.04 (±0.70)	483.75 (±0.14)	−10.5 (±0.73)	431.16 (±0.56)
August	−9.31 (±0.70)	413.98 (±0.33)	−11.12 (±0.74)	451.10 (±0.45)	−10.3 (±0.69)	420.76 (±0.16)	-----------	----------------

presents the measured values and the related uncertainties based on the repeatability of the individual measurements.

A Mann–Whitney test was applied to assess whether, within the same area, the δ¹³C(CO₂) medians of the different sites were statistically equal. For Cigognola and Santa Maria della Versa, the p-value was ≃0.66, while for Illasi and Mezzane it was ≃0.77. These results indicate that within each respective area, the δ¹³C(CO₂) values are statistically indistinguishable. Therefore, it was appropriate to calculate a weighted mean and the associated standard deviation for each area (standard deviation was computed accounting for the different weights used in the mean).

$$\mathit{Cigognola}\mathit{and}\mathit{Santa}\mathit{Maria}\mathit{della}\mathit{Versa}: {\mathsf{\delta }}^{13}{\mathrm{C}}_{\mathrm{C}\mathrm{O}2/\mathrm{V}\mathrm{P}\mathrm{D}\mathrm{B}}=\left(-9.67\pm 0.74\right)\mathrm{‰}$$

(19)

$$\mathit{Illasi}–\mathit{Mezzane}: {\mathsf{\delta }}^{13}{\mathrm{C}}_{\mathrm{C}\mathrm{O}2/\mathrm{V}\mathrm{P}\mathrm{D}\mathrm{B}}=\left(-11.12\pm 0.72\right)\mathrm{‰}$$

(20)

Subsequently, we tested whether the medians of the two areas were statistically equivalent. The Mann–Whitney median test yielded a p-value of approximately 0.05, indicating a low probability that the medians were equal.

Using Equation (6), the mean carbon isotope enrichment factor between CO₂ and sugar ${(\mathsf{\epsilon }}_{\mathrm{C}{\mathrm{O}}_2-\mathrm{s}\mathrm{u}\mathrm{g}\mathrm{a}\mathrm{r}}$) and its standard deviation were calculated for Cigognola and Santa Maria della Versa (18.0 ±0.8‰) and Illasi–Mezzane, (14.3 ± 0.8‰) (

Table 2. Values of CO₂—sugar enrichment factor’s (ε in ‰) at Cigognola, Santa Maria della Versa and Illasi–Mezzane. The values are listed in the same order of the date of harvest, and, in brackets, the data of the harvest are indicated.

Cigognola	Santa Maria Della Versa	Illasi–Mezzane
18.5 (26 August 2022)	18.1 (26 August 2022)	14.3 (25 August 2022)
17.9 (5 September 2022)	17.4 (5 September 2022)	13.8 (26 August 2022)
19.3 (12 September 2022)	18.0 (12 September 2022)	13.9 (30 August 2022)
18.1 (19 August 2022)	16.6 (19 August 2022)	13.5 (2 September 2022)
		15.2 (5 September 2022)
		13.4 (6 September 2022)
		16.0 (4 October 2022)
		14.1 (18 October 2022)
		14.7 (18 October 2022)

). The Mann–Whitney test revealed that the average enrichment factors of the two areas were significantly different (p < 0.001).

Since the enrichment factor (ε) links the δ¹³C of sugar to the δ¹³C (CO₂) values, we investigated whether a correlation exists between ε values and either the hydroclimatic balance (BIC) or cumulative rainfall, both of which were previously used to assess the relationship with the δ¹³C (sugar) values (Figure 3, Figure 4, Figure 5 and Figure 6). This analysis was conducted only for the localities where atmospheric CO₂ was sampled (Cigognola, Santa Maria della Versa and Illasi–Mezzane). A significant linear correlation was found only between ε and cumulative rainfall (Figure 7).

Figure 7 supports the existence of a relationship between cumulative rainfall, atmospheric δ¹³C (CO₂), and the δ¹³C of sugar. Specifically, cumulative rainfall accounts for approximately 60% of the variability in the enrichment factors. Therefore, we can conclude that both rainfall (mm) and atmospheric δ¹³C (CO₂) are the major drivers influencing the carbon isotopic composition of grape sugar. To identify which parameter has the greatest influences on δ¹³C of sugar, we plotted δ¹³C (sugar) against cumulative rainfall in Figure 8, using the same rainfall data presented in Figure 7.

• The coefficient of determination in Figure 8 is slightly higher than that in Figure 7. This suggests that the most important factor influencing the carbon isotopic composition of sugar is the amount of water absorbed by the plant, rather than the isotopic composition of atmospheric CO₂, particularly when water is available in substantial quantities. This finding further supports the hypothesis that water is most likely the limiting reagent in photosynthesis reaction (Equation (4)). It is also noteworthy that the weighted mean of atmospheric δ¹³C (CO₂) at the Illasi–Mezzane sites is more negative than that measured at Cigognola and Santa Maria della Versa. Despite this, the mean δ¹³C value of sugar at Illasi–Mezzane (δ¹³C_mean = −25.07 ± 0.80‰) is higher than that observed at Cigognola and Santa Maria della Versa (δ¹³C_mean = −27.17 ± 0.74‰). This difference is likely due to the lower cumulative rainfall at the Illasi–Mezzane sites, where total precipitation was less than 200 mm.

4. Conclusions

This study investigated the influence of climatic/meteorological variables on the carbon isotopic composition of grape must sugar. The main findings from both study areas are the following:

1

For the 2022 harvest, a clear relationship was observed between the δ¹³C values, cumulative rainfall, and BICc, indicating that increased water uptake by plants corresponds to lower δ¹³C values in grape sugar.

2

At Cigognola, Santa Maria della Versa, and Illasi–Mezzane sites, a strong correlation was found between the enrichment factor (ε) and cumulative rainfall. This relationship altogether considers the δ¹³C value of atmospheric CO₂, sugar, and plant water uptake, highlighting that water availability is the primary influencing factor.

3

A slightly stronger correlation was observed between cumulative rainfall and the δ¹³C values of must sugar compared with the correlation with the enrichment factor (ε). This suggests that water availability exerts a more dominant influence on the carbon isotopic signature than either the isotopic composition or the partial pressure of atmospheric CO₂. Grape must δ¹³C appears to primarily reflect the plant’s water status

4

To strengthen the findings of this preliminary study, it is necessary to replicate the research over the coming years (ideally for at least ten years) and in different regions of Italy. This to assess whether BICc and water availability consistently show a correlation with the carbon isotopic composition of grape sugar. If this is not the case, it will be important to identify which parameters, beyond actual vapor pressure or vapor pressure deficit (VPD), may have a stronger influence on the isotopic composition of grape sugar. Furthermore, to confirm whether water uptake by plants plays a more significant role than the isotopic composition of atmospheric CO₂, daily air sampling would be required to obtain more accurate estimates of both the concentration and the isotopic signature of atmospheric CO₂. Finally, we emphasize the importance of collecting enough samples per site to ensure good reliability of the results.