Abstract
Red Cesanese grapes were harvested at 21 Brix. Berries were destemmed manually and placed in perforated plastic trays inside a small ventilated tunnel at 10°C with 1.5 m/sec of air flow, 10°C with 2.5 m/sec of air flow, and 20°C with 1.5 m/sec of airflow as a control. Relative humidity was 45%. The experiment was stopped at 20% berry weight loss, after 26, 22, and 16 days for 10°C at 1.5 m/sec, 10°C at 2.5 m/sec, and 20°C at 1.5 m/sec, respectively. Sugar content rose to 24–25 Brix. Carbon dioxide production from chilled berries under a higher air flow rate was constantly higher (~30%) than at a lower ventilation, and similar to that of the 20°C sample. Berries lost firmness (higher deformation), increased hue angle, and decreased chroma, regardless of treatment. Berries kept at a higher flow rate had magnetic resonance images similar to those of the berries kept at 20°C, with diffuse dark areas in the mesocarp. Analysis of alcohol dehydrogenase in the direction of the oxidation of ethanol to acetaldehyde revealed significantly higher activity at 20°C than at 10°C at 20% of weight loss, and ethanol was lower. The fluorescence (Fα) pattern confirmed a different stress rate depending on temperature and flow rate. Near-infrared–acousto-optic tunable filter (NIR–AOTF) analysis revealed a different absorbance level at a specific wavelength range depending on the treatment and the rate of weight loss. A significant decrease in polyphenols occurred in 10°C samples. Flavonols and stilbenes increased significantly at 20°C, confirming a supposed higher rate of water stress at 20°C.
Water stress begins when tissue moisture content, intracellular or extracellular, deviates from the optimum and the turgor pressure of the cell drops below its maximum value (Kays 1997). Skin cells are the most sensitive to changes in water potential and are rich in secondary metabolites with defense activity, which are mainly synthesized via the phenylpropanoid pathway. In grapes, field water deficit has been observed to increase anthocyanins in Cabernet franc (Matthews and Anderson 1988) and Shiraz (Ginestar et al. 1998) and anthocyanins and skin tannins in Cabernet Sauvignon grapes (Roby et al. 2004, Kennedy et al. 2002). Moreover, water deficit stimulated sugar accumulation (Roby et al. 2004, Kennedy et al. 2002).
In Italy, most sweet wines (as well as some dry wines such as Amarone and Sfurzat in northern Italy) are produced from grapes dried in an open or closed environment. Modification of the drying environment greatly affects grape metabolism, and thus the final quality of the wine (Barbanti et al. 2008). Accurate control of temperature, relative humidity (RH), and air flow significantly affects grape metabolism in terms of volatile compounds (Chkaiban et al. 2007). Temperature plays a key role in accelerating or delaying the desired water loss during the dehydration of winegrapes, but it is mainly important for the modulation of volatile compound metabolism and the formation of volatile acidity (Mencarelli et al. 2006). Alcohol dehydrogenase (ADH) activity and anaerobic metabolites such as ethanol, acetaldehyde, and ethyl acetate increase significantly when berry water loss reaches ~20% under closed, controlled, environmental conditions (Costantini et al. 2006), but this weight loss threshold is reached earlier when the vapor pressure deficit (VPD) changes continuously under noncontrolled drying conditions (Chkaiban et al. 2007).
For red wine obtained from dehydrated grapes, temperature management is even more important to maintain phenol fractions, in addition to the control of volatile acidity. Drying winegrapes under the sun in a hot climate, as in Jerez de la Frontera and Montilla-Moriles (Spain) or in Sicily (Italy), completely changes the aroma panorama of grape musts, with an increase in 5-methylfurfural, furfural, and anaerobic volatiles such as ethanol, phenylethanol, acetoin, ethylacetate, isoamyl alcohol, and γ-butyrolactone (Franco et al. 2004). In large grape drying facilities (500 metric tons of grapes) where it is difficult to maintain uniform temperature and relative humidity, air is circulated across the grape pallet. Air speed plays an important role in the release of water from berries and can modify the rate of water stress, whatever the temperature (Mencarelli et al. 2006).
Commercially, the use of destructive analyses to follow the progress of grape dehydration is expensive and time consuming. Current nondestructive techniques can be a powerful tool, both scientifically and commercially. Magnetic resonance imaging (MRI) has been used on grapes to monitor structural characteristics and ripening progress (Andaur et al. 2004). In kiwifruit, MRI has been used to study water loss. Measurements from serial MR images revealed fruit volume decreases and relaxation measurements [spin-lattice (T1) and spin-spin (T2)] indicated that there were significant differences between the localized aqueous environments of the kiwi core, inner pericarp, and outer pericarp and that water content (proton density) differed across the fruit (Burdon and Clark 2001). Visible near-infrared (vis-NIR) spectroscopy is an established technique for measuring the concentration of chemical constituents in agricultural products. The NIR region contains information concerning the relative proportions of C-H, N-H, and O-H bonds, which are the primary structural components of organic molecules. Recently, particular interest in using NIR on grapes and wine has arisen for the analyses of phenolics (Cozzolino et al. 2004) and glycosylated compounds involved in grape juice aroma (Cynkar et al. 2007). Moreover, the use of NIR-acousto-optic tunable filter (AOTF) (Gonzaga Barbieri and Pasquini 2005) has permitted the zonation of Cabernet Sauvignon based on analyses of total phenols, total anthocyanins, malvidin, malic and tartaric acids, and sugars (Kaye and Wample 2005). Weight loss discrimination during grape drying has been performed using Fourier transform IR (FTIR) (Femenia et al. 1998).
In this study on the partial dehydration of Cesanese red grapes, two air flows at 10°C were compared to one air flow at 20°C as a control, with RH kept at 45%. In addition to changes in the physical parameters of the berries such as color, berry deformation, and water loss, particular attention was placed on the change in specific nonflavonoid and flavonoid compounds and on anaerobic metabolism (alcohol dehydrogenase and relative volatile compounds). In addition, nondestructive techniques such as MRI, NIR–AOTF, and the HarvestWatch fluorescence sensor were used to further elucidate physical and chemical changes.
Materials and Methods
Experimental procedure and quality analyses.
Red grapes (Vitis vinifera L. var. Cesanese) were carefully harvested at 21 Brix in an area near Rome where this variety is widely grown and sorted for uniform berry size and soundness. Berries were detached from clusters by cutting the pedicel 3–4 mm from the berry-pedicel connection using scissors. Berries were placed in perforated boxes (60 x 40 x 15 cm) in a single layer. For each test, two perforated boxes with 1000 berries each were placed in a small metallic tunnel (45 x 45 x 100 cm) adapted with an exhaust fan with air-flow regulation. The small tunnels were placed in a thermohygrometric controlled room (12 m3). Two tunnels with different air flows, 1.5 and 2.5 m/sec, respectively, were set at 10 (±1)°C, while another tunnel with an air flow of 1.5 m/sec was set at 20 (±1)°C (considered the control, as it is the dehydration condition used most often for grapes in a closed system). An air speed of 1.5 m/sec among grape trays is typically used in commercial dehydration rooms, operating without mechanical control of the temperature, but using fans for outside air intake during fall and winter. An air speed of 2.5 m/sec is proposed to compensate for the low vapor pressure deficit (VPD) at 10°C without modifying temperature and relative humidity (RH) of the environment. The dehydration duration is important commercially, and low temperature significantly prolongs dehydration time. Thus, to accelerate the process while maintaining the low drying temperature, we compared two air flows at a lower temperature. The relative humidity was set at 45% (±5%) in both rooms. Room fans were regulated at very low air flow just sufficient to maintain uniform inside temperature.
Each individual berry was initially measured for height, width, and weight; during the test, individual berries were weighed daily to calculate weight loss. After loss of 5, 10, 15, and 20% of the original weight, 50 berries were sampled from each treatment and analyzed immediately for color parameters with a Minolta C2500 Spectrophotometer (Konica Minolta, Ramsey, NY) on two opposite sides of the berry; values are reported as hue angle (arctan b/a) and chroma (a2 + b2)½. Whole berry firmness was monitored using a deformation test under nondestructive force. The berries were placed widthwise over a plate on an Instron Universal Testing Machine (model 4301; Instron Inc., Canton, MA), analyzed for firmness, and compressed with a flat compression anvil (55 mm diam) to a fixed load of 1 N, at a rate of 25 mm min−1. Data were expressed in terms of millimeters of deformation under a force of 1 N. Soluble solids content (SSC) of the juice obtained from the berries was measured using a digital refractometer (Atago, Tokyo, Japan). Respiration as CO2 production was monitored with an Oxycarb infrared analyzer (Isolcell, Bozen, Italy) by placing 25 berries inside a 0.5-L glass jar with a lid adapted with a rubber stopper for insertion of analyzer needles at the time of analysis. Three jars were averaged for each treatment. Berries were sorted and marked with a pen at the beginning of the experiment; at each sampling time, they were collected from the tunnels, enclosed in the jars, and closed with the lid for 1 hr at the same temperature as drying. After this interval, an air sample was taken from the jar headspace and injected into the analyzer for CO2 analysis.
Nondestructive measurements.
MRI measurements were performed with an Avance 300 MHz spectrometer (Bruker Biospin, Milan, Italy) equipped with a cylindrical birdcage single-tuned nucleus (1H) coil probehead with an inside diameter of 60.0 mm. The water signal was monitored and used for image reconstruction. Gradient echo (GEFI) and multi-slice–multi-echo (MSME) experiments, m_gefi_ortho, m_msme_ortho, m_rare64 and m_ge3d, respectively (Bruker library), were performed according to standard procedures. In GEFI measurements, which generate echoes only by applying gradient pulses, the field of view was 25.0 mm x 25.0 mm, matrix size was 128 x 128 pixels, and spectral width was 100.0 kHz. Echo and repetition times were set to 2.445 msec and 60.0 msec, respectively. The number of scans was one; slice thickness was 1.0 mm; and excitation pulse was a sinc3. The data were processed to obtain images of 128 x 128 pixels and a field of view of 25.0 mm x 25.0 mm. The processing mode was FT_MODE, which was complex_FFT, and spike elimination was allowed. In MSME experiments, which produce echoes via a spin echo-based sequence and give T2-weighted images, field of view was 25.0 mm x 25.0 mm, matrix size was 128 x 128 pixels, spectral width was 100.0 kHz, echo and repetition times were set to 12.5 msec and 4000.0 msec, respectively, number of echoes and images was 24, number of scans and dummy scans was one, slice thickness was 1.0 mm, and excitation pulse was a sinc3. The data were processed to obtain images of 128 x 128 pixels, and a field of view of 25 mm x 25.0 mm. The processing mode was FT_MODE, which was complex_ FFT, with spike elimination allowed.
During the dehydration process, chlorophyll fluorescence of the grape tissue was monitored using a Harvest-Watch system (Isolcell, Laives, Bozen, Italy), which has been effective at detecting low oxygen tolerance limits in fruits and vegetables (Prange et al. 2003). In our case, the goal was to determine the ability to detect metabolic changes due to water loss. Data are reported as raw Fα, which is an estimation of the F0. Berries from ventilated tunnels were sampled after 5, 10, 15, and 20% weight loss and subdivided into two sets (25 berries each); each set was placed for 4 hr inside a HarvestWatch box with the sensor facing the berries. Spectral detection was conducted on the same 50 grape berries using a Luminar 5030 Miniature Hand-held NIR Analyzer (Brimrose Corporation, Baltimore, MD), which is based on the AOTF-NIR principle. Two different measurements were performed on each intact grape berry by contact between the external gun of the NIR device and the epicarp of the fruit, using the diffuse reflectance method of detection. Detection was conducted in the 1100–2300 nm range, with 2 nm wavelength increments and 10 spectra per average, which represented a single measurement. The average of the two measurements was the spectral response of the berry.
Raw spectra were statistically pretreated for absorbance (log 1/R) transformation and first derivative (with 9 points of smoothing) using SNAP! 2.03 software (Brimrose). Pretreated data were used for principal component analysis (PCA) and partial least square (PLS) calculations, which were performed using Unscrambler v9.2 software (CAMO ASA, Oslo, Norway). Full cross-validation was used and the statistical indexes R2, RMSEP, and Bias were used to determine the significance of the calculations.
Biochemical and chemical analyses.
Alcohol dehydrogenase (ADH) enzyme analysis was performed as described elsewhere (Costantini et al. 2006). Volatile compounds were measured using solid-phase microextraction (SPME), following a published procedure (Chkaiban et al. 2007). A total of 5 mL grape berry juice was transferred to a 25-mL glass miniflask (Supelco, Sigma-Aldrich, St. Louis, MO) containing a small Teflon-coated stirring bar with a screwtop and PTFE-faced silicone septum, to which 5 mL saturated CaCl2 (1:1 w/v) was added. The mixture was homogenized with 200 μL 1-penten-3-one (5 g/L in Milli-Q water) as a standard. The solution was kept under continuous stirring in a thermostatic bath at 20 ± 2°C. After 10-min equilibration, volatiles from the juice headspace were extracted for 30 min using a 100- μm PDMS SPME fiber (Supelco Inc., Bellafonte, PA). Before each exposure, the fiber was cleaned in a 250°C injection port for 7 min.
After extraction, the SPME fiber was transferred to the injection port and thermally desorbed at 230°C for 7 min in a splitless mode. Gas chromatography (GC) analyses were conducted using a Trace GC Ultra (Thermo Fisher Scientific, Waltham, MA) equipped with a 60 m x 0.25 mm x 0.25 μm DB-Wax column (J&W Scientific, Folsom, CA). Helium was used as the carrier gas (27 cm/s). The oven temperature was maintained at 40°C for 7 min and then programmed to reach 230°C at a rate of 3°C/min, with a final isotherm of 30 min, using a high-sensitivity flame ionization detector (FID) at 260°C. Compound identification was achieved using a Shimadzu 17A GC–MS and a QP 5050A MS (Shimadzu, Kyoto, Japan) and matching spectra against the NIST 107 and NIST 21 libraries and by matching GC retention times against standards. Results were expressed as peak area x 1000.
Polyphenols were identified and quantified by high-performance liquid chromatography–electrospray interface –mass spectrometry (HPLC–ESI–MS). Frozen grape berries were powdered in liquid nitrogen and, after the seeds were removed, lyophilized. A weighed amount of the lyophilized sample was quantitatively extracted with an 8:2 (v/v) methanol-ethanol mixture at room temperature for 2 hr in a round-bottom flask. The extraction was repeated twice and the collected supernatants were concentrated in a rotavapor within a 35°C water bath. The residue was quantitatively recovered in 1 mL 8:2 (v/v) methanol-water and used for HPLC– ESI–MS identification and quantification of polyphenols. The experiments were performed with a Shimadzu HPLC–MS LCMS-2010 unit, comprising a SCL-10Avp system controller, two LC-10ADvp Solvent Deliver y Module pumps, a SPD-M10Avp UV-vis Photodiode Array Detector, a single quadrupole 2010 mass analyzer equipped with an ESI, with nitrogen as the nebulizing and drying gas. MS acquisition was performed with the ESI in the negative ionization mode. Optimized conditions were determined by flow injection analysis (FIA) of standard solutions of the analytes at three different concentrations ranging from 0.1 to 50 mg/L. System control and data processing were carried out by Shimadzu LCMS solution software running on a personal computer. The compounds were separated using a Polaris C18A column (150 x 2 mm i.d., 5 μm) (Varian Inc., Lake Forest, CA) in conjunction with a C18 (30 x 2 mm, 5 μm) guard cartridge column; column temperature was 30° ± 1°C. Separations were performed by a multi-step gradient of increasing concentration of acetonitrile in an acetonitrile-water mixture containing 5% (v/v) formic acid, at a flow rate of 0.2 mL/min. Samples were introduced into the column by a semi-microinjection valve (model 8125; Rheodyne, Cotati, CA) with a 5-μL sample loop. Column effluent was passed through the PDA detector before being directed to the quadrupole MS–ESI.
Results and Discussion
Twenty percent of weight loss, which was the fixed final dehydration, was reached in 26, 22, and 16 days for 10°C at 1.5 m/sec air flow, 10°C at 2.5 m/sec, and 20°C at 1.5 m/sec, respectively (Figure 1⇓). The R2 of the straight regression line between weight loss and time was 0.99 for all samples. Higher ventilation (2.5 m/sec) at 10°C significantly increased the water loss rate (shorter time to reach 20% weight loss) compared with slower air flow (1.5 m/sec), but did not reach the rate of the 20°C sample. The pick-up efficiency (capacity to absorb moisture) of 1 m3 of air at 10°C and 45% RH is 2.5 g of water vapor versus 3.3 g at 20°C (Rozis 1997). When the berry is turgid, the vapor pressure deficit (VPD) at 10°C is 8–9 mbar and rises to 13–14 mbar at 20°C. Thus, the escape rate of water vapor from the tissue at 20°C is almost twice that at 10°C. To accelerate the berry water loss at 10°C, the relative humidity can be reduced, but for large commercial facilities that is very costly. The optimal solution is to increase air flow over the berries. By increasing air flow at 10°C, water loss accelerates because there is a faster removal of the boundary layer from the berry peel surface and continuous formation of VPD. Sugar content increases proportionally to berry water loss up to 24, 24, and 25 Brix for 10°C at 1.5 m/sec, 10°C at 2.5 m/sec, and 20°C at 1.5 m/sec, respectively.
Rate of weight loss of grape berries at 10°C, 45% RH, and 1.5 and 2.5 m/sec air flow, and at 20°C, 45% RH, and 1.5 m/sec air flow. Data are the mean of 50 berries. Bars indicate SD.
In terms of physiology, ventilation affected respiration, which was constantly higher (~30%) at 10°C in berries kept under a higher flow rate (2.5 m/sec) (Figure 2⇓). At the end of the experiment, CO2 production was 2.4 and 1.8 mL/(kg·h), for 2.5 and 1.5 m/sec samples, respectively. At 20°C, respiration was slightly higher, not significantly different than the 10°C at 2.5 m/sec sample, until 10% weight loss (wl) was reached; at 15% wl, CO2 production was much higher (4 mL/(kg·h)) and then decreased, at 20% wl, to the same level as the 10°C at 2.5 m/sec sample. The increase in respiration in plant cells undergoing water loss has a double effect: first, it produces metabolites such as proline, useful for protecting against water loss (termed “compatible”), and, second, it may lead to the production of potentially dangerous reactive oxygen species (Hoekstra 2005). At 10°C, berries kept at 1.5 m/sec flow rate showed ~25% lower CO2 production during the entire experimental period, which is important in terms of CO2 release in the atmosphere and less heat produced (21.4 J/(kg·h) vs. 26.7 J/(kg·h)), particularly when taking into account that some dehydration plants for Amarone wine can stack 500 metric tons of grapes. Aside from this technical aspect, the higher respiration rate means more substrate consumed, with consequences in terms of grape quality.
Rate of carbon dioxide production at 10°C, 45% RH, and 1.5 and 2.5 m/sec air flow, and at 20°C, 45% RH, and 1.5 m/sec air flow. Data are the mean of three jars, each containing 25 berries. Bars indicate SD.
For physical characteristics such as color, berries of all samples showed a significant increase in hue angle and a decrease in chroma, regardless of the dehydration treatment (Table 1⇓), mainly because of the increase in “a” (redness) and the decrease in “b” (blueness). Berry firmness measured as berry deformation (mm) was ~1 mm at 5% wl and rose to ~2 mm in all samples, without significant difference, at 20% (Table 1⇓). Thus, the weight loss percentage significantly affected berry deformation, but the weight loss rate (weight loss/day) did not.
Hue angle, chroma, and deformation under a force of 1 N of berries at indicated conditions. Relative humidity was constant at 45% in all samples; 0, 5, and 20% indicate berry water loss.
To better understand whether or not the effect of weight loss rate was really negligible, texture changes occurring during dehydration were monitored by MRI (Figure 3⇓). No significant difference was observed between the images taken at time 0 (initial time) and at the time of 5% wl. At 10% wl, berries maintained at 20°C began to show some black areas (mobile water) under the skin tissue. At 15% wl, these black areas became larger at 20°C and began to appear under the skin in samples at 10°C. At 20% wl, 20°C berries were almost completely black, 10°C at 2.5 m/sec berries showed large areas under the skin extending toward the area containing the seeds, and the tissue of 10°C at 1.5 m/sec berries was less disorganized. Thus, notwithstanding the same percentage of weight loss, the rate of dehydration affects tissue integrity and, consequently, the amount of stress. This conclusion is consistent with the CO2 production rate data (Figure 2⇑), where berries at 10°C with faster air flow produced more CO2 and berries at 20°C had even higher values.
Magnetic resonance image of berries at the beginning of the experiment (A), dehydrated at 10°C at 1.5 m/sec (B), at 10°C at 2.5 m/ sec (C), and at 20°C at 1.5 m/sec (D). Each two pictures in the B, C, and D sequences, from top to bottom, refer to 5, 10, 15, and 20% of berry weight loss. Analyses were performed on 10 berries for each sample and each sampling time. For clarity, two berries for each sample and at each sampling time are shown.
This stress effect was further confirmed by the HarvestWatch curves (Figure 4⇓), where there was a decline in chlorophyll fluorescence (Fα), which appeared more rapid in berries kept at 20°C and 10°C at 2.5 m/sec. HarvestWatch has been proposed to measure the anaerobic compensation point (ACP) or the fermentation threshold (FT) during the ultra-low oxygen storage of apples, since the reduced oxygen-inducing cytoplasmic acidosis causes an increase in fluorescence Fα from chloroplasts (Prange et al. 2005). Adenosine 5′-triphosphate (ATP) production and consumption must be balanced. If there is either too little production or too much consumption, then acidosis will occur and there will be an increase in Fα. Any abiotic stress will reduce ATP production via aerobic respiration and force the cell to rely more on anaerobic respiration. In contrast, during berry water loss, fluorescence Fα decreased, and the rate of decrease was lower for the sample at 10°C at 1.5 m/sec, which was the soundest under the MRI test. The stability or slight decrease in values in the 10°C at 1.5 m/sec sample compared with the other two samples (10°C at 2.5 m/sec and 20°C at 1.5 m/ sec) during dehydration would indicate a good ATP situation (ATP/adenosine 5′-diphosphate [ADP ratio]) and no acidosis occurring. Water stress reduces photochemical efficiency in plants (Cai et al. 2007), including grapevine (Smithyman et al. 2001).
HarvestWatch readings in relation to weight loss percentage (Fα: chlorophyll fluorescence). Two sets (25 berries each) of berries from each ventilated tunnel kept at the indicated conditions, at the indicated percentage of weight loss, were placed in the HarvestWatch box and continuously measured for 4 hr. Reported results are the interpolated data from daily measurements plotted versus grape weight loss.
Analysis of ADH in the direction of ethanol oxidation, as an index of anaerobic shifting during grape drying (Costantini et al. 2006, Chkaiban et al. 2007), and related metabolites (ethanol, acetaldehyde, and acetic acid), revealed significantly higher activity at 20°C than at 10°C after 20% wl; in contrast, at 20°C, ethanol was lower than in the other two samples but higher than in the sample measured at the initial time, indicating an anaerobic process (Table 2⇓). The most ethanol and the lowest ADH activity were found in the sample at 10°C at 1.5 m/sec. Acetic acid was significantly high at the initial time, then decreased in the 10°C samples, while it increased slightly at 20°C. Acetaldehyde was most abundant at 20°C, while ethylacetate was least abundant. It appears that the anaerobic process took over in all samples because of dehydration, with accompanying formation of ethanol and acetaldehyde. At 20°C, the significant increase in ADH, the low ethanol concentration, and the higher acetaldehyde and acetic acid content would confirm an oxidation of ethanol to acetaldehyde to acetic acid, as seen in white grapes under dehydration (Chkaiban et al. 2007) and as explained elsewhere (Perata and Alpi 1993). A higher pH favoring ADH activity in the reverse direction could be provided by protein degradation, thus a higher formation of NH4+ cation, and by salification of acids with Ca++ and K+, which increase significantly during grape cell dehydration (Di Mambro et al. 2008). The increase in pH is a natural physiological behavior during grape ripening following berry softening, and thus during cell wall degradation (Barnavon et al. 2000); this behavior also occurs during berry dehydration and the pH increases as a result of the decrease in free acids (Franco et al. 2004). In contrast, at 10°C, anaerobic metabolism still appears to be moving in the direction of acetaldehyde reduction to ethanol. In this case the increased formation of ethylacetate and other acetate esters, especially iso-amylacetate, would be justified by the depletion of acetic acid and the high availability of ethanol (Table 2⇓). An oxidative process seems to be confirmed by the increase in C6 compounds (E-hex-3-en-1-ol, Z-hex-3-en-1-ol, Ehex- 2-en-1-ol, E-hex-2-enal) occurring in samples kept at 20°C (Table 2⇓).
Alcohol dehydrogenase (ADH) activity (nmoles/g d.w.), ethanol, acetaldehyde, acetic acid, ethylacetate, isoamyl, and C6 content of grapes kept at the indicated conditions at 20% weight loss.
The application of NIR-AOTF revealed a significant difference in absorbance among samples. High absorbance peaks were revealed at 1120, 1380, 1890, and 2170 nm; lower absorbance peaks at 1350, 1750, and 2300 nm; and negative peaks at 1230 and 1519 nm (Figure 5⇓). A difference in magnitude at each wavelength was observed in relationship to water loss (5, 10, 15, and 20%) but absorbance at each weight loss stage differed depending on the samples. The peaks at 1380 and 1980 corresponding to the water overtones (Cozzolino et al. 2004) indicate that at 10°C at 1.5 m/sec, absorbance in the 1300–1400 nm region increased linearly, from 5 up to 20% wl; at 10°C at 2.5 m/sec, there was a larger gap between 5 and 10% wl and 15 and 20% wl. At 20°C, this gap increased and absorbance at 5 and 10% wl overlapped. During dehydration, the acute absorption band of water at 1406 nm decreases in intensity and an absorption band at 1430 nm begins to increase (Büning-Pfaue 2003). This finding indicates a regular water loss at 10°C, especially at lower air flow, and confirms another report (Wang and Brennan 1995), which found that shrinkage occurs first at the surface and then gradually moves to the internal tissue as drying time increases. At a slow drying rate, the moisture content at the center of a grape is not much higher than at the surface, the internal stresses are minimized, and the material shrinks fully onto a solid core. Weight loss prediction by means of absorbance data measured with NIR gave an R2 of 0.93, with an RM-SEP of 2.52, and a Bias of −0.0048. The coefficient of determination (R2) between NIR absorbance and tissue deformation (mm) was 0.80, with an RMSEP of 0.24 and a Bias of −0.0001.
Near-infrared absorbance spectra of berries at 10°C at 1.5 m/ sec (A), 10°C at 2.5 m/sec (B), and 20°C at 1.5 m/sec (C). The four lines refer to the sampling time (weight loss): T1 = 5%, T2 = 10%, T3 = 15%, T4 = 20%. Data are the mean of 50 berry readings, two readings per berry.
Polyphenol analyses revealed significant changes in single compounds, depending on treatment. The overall concentration was higher at the initial time, and a significant decrease occurred at 20% wl, from 4092 mg/ kg d.w. to ~3500, 3400, and 2900 for 20°C at 1.5 m/sec, 10°C at 2.5 m/sec, and 10°C at 1.5 m/sec, respectively. Among the hydroxycinnamic acids, caftaric acid is the most important and one of the most sensitive to atmosphere variation, particularly in relationship to water stress (Karadeniz et al. 2000) or UV light (Benítez et al. 2003). It decreased by ~43, 21, and 39% at 20% wl for 10°C at 1.5 m/sec, 10°C at 2.5 m/sec, and 20°C at 1.5 m/ sec samples, respectively (Table 3⇓). Loss of caftaric and coutaric acids was observed in sundried Thompson seed-less grapes, and this loss was reduced by the use of SO2 (Karadeniz et al. 2000). Red Barbera grapes dried under controlled ventilation at room temperature or at external climate temperature lost anthocyanins, hydroxycinnamic acids, and flavan-3-ols to a significant degree (Borsa and Di Stefano 2000). In Aleatico grapes dehydrated up to 26% wl, trans-caftaric acid and total polyphenols (mg gallic acid/1000 berries) increased and remained constant, respectively, when the temperature was between 17 and 23°C, but declined significantly, especially the total polyphenols, when the temperature was increased to 30°C (Frangipane et al. 2007).
Specific polyphenols (mg/kg d.w.) of grapes dried at the indicated conditions at initial time and at 20% weight loss.
Catechin decreased from 331.67 mg/kg by 30, 22, and 14% at 20% wl for 10°C at 1.5 m/sec, 10°C at 2.5 m/ sec, and 20°C at 1.5 m/sec samples, respectively (Table 3⇑). In contrast, epicatechin increased significantly. In Pinot noir grapes dehydrated to 14–16% wl, at 22°C and 35% RH plus ventilation, proanthocyanidins per berry decreased, as did the degree of polymerization (Moreno et al. 2008). Flavonols such as quercetin and kaempferol increased significantly in all samples at 20% wl, especially in the 20°C sample, while myricetin decreased without a significant difference among samples (Table 3⇑). Among the anthocyanins, malvidin-3-glucoside diminished significantly in all samples from 1841.62 mg/kg d.w. by ~24, 14, and 17% at 20% wl for 10°C at 1.5 m/sec, 10°C at 2.5 m/sec, and 20°C at 1.5 m/sec samples, respectively. Peonidin-3-glucoside diminished significantly in samples at 10°C, but remained stable in samples at 20°C. Delphinidin-3-glucoside decreased in all samples but significantly only in samples with 1.5 m/sec flow rate regardless of temperature; cyanidin decreased in 10°C samples. A decrease of approximately 40% in anthocyanins following dehydration to 30% wl was observed in Cesanese grapes dehydrated under controlled conditions with temperature starting at 10°C and rising to 25°C (Pietromarchi et al. 2007). Another study found no significant changes in anthocyanins, but weight loss was lower (14–16%) (Moreno et al. 2008). trans-Resveratrol remained constant or decreased slightly in 10°C samples, but increased significantly in 20°C samples, while trans-piceid decreased significantly at 10°C, but rose considerably at 20°C (Table 3⇑). High temperature (45°C for 36 hr) and a long wilting time (94 days) was very effective for increasing stilbene compounds (resveratrol and trans-piceid) and the expression of stilbene synthase (Versari et al. 2001). Thus, it appeared that at 20°C, the loss of flavonols (kaempferol and quercetin), catechin, and anthocyanins was lower than at 10°C, and stilbenes increased considerably.
Conclusions
Reducing the temperature from 20°C to 10°C for the partial dehydration (20%) of Cesanese grapes for wine production requires a longer time but induces a lower degree of texture stress and, consequently, lower physiological stress in the berries. At 20°C, higher ADH activity (ethanol to acetaldehyde), lower ethanol, and higher acetaldehyde and acetic acid indicated an advanced anaerobic process. An increase in air flow to 2.5 m/sec at 10°C accelerates water loss but causes texture changes similar to 20°C, although less dramatic, as shown by MRI. Harvest-Watch readings confirmed the higher stress in samples at 20°C and at 10°C with a higher flow rate. NIR spectra revealed a more regular change in absorbance in berries kept at 10°C with a lower flow rate. The lower decrease in polyphenols (anthocyanins) or the increase, such as for stilbenes and flavonols at 20°C, seems to confirm the hypothesis of a higher rate of water stress, which can be regarded as positive for red grape varieties from a commercial viewpoint. Finally, nondestructive techniques are an efficient tool for following the chemical and physical progress of grape dehydration.
Footnotes
-
Acknowledgments: The authors thank R.K. Prange and J.M. DeLong for experimental support and their suggestions in the use of HarvestWatch, as well as Roberto Forniti for technical assistance.
-
This research was supported by the Italian Ministry of Agriculture (MIPAAF), MUVON Project.
- Received April 2008.
- Revision received August 2008.
- Accepted November 2008.
- Copyright © 2009 by the American Society for Enology and Viticulture
Literature Cited
Vol 60 Issue 1
