Abstract
Background and goals Interspecific hybrid grapes are commonly used for wine production in areas where environmental pressures inhibit traditional Vitis vinifera cultivation. However, red hybrid grapes have diverse anthocyanin profiles, and techniques that optimize color in red V. vinifera wines are often ineffective in hybrid wine production. Because the chemistry of hybrid grape anthocyanins is largely unknown, the reactions they undergo during ripening, wine production, and aging are poorly understood. To characterize the anthocyanin composition of three unstudied cultivars and to clarify the effect of vine microclimate, skin anthocyanin profiles were assessed for shaded and unshaded fruit from regionally grown cool-climate hybrid grapes (Vitis spp.) Corot noir, Maréchal Foch, and Marquette.
Methods and key findings Berry samples were collected throughout ripening from triplicate blocks of exposed and shaded fruit and skin anthocyanins were extracted and characterized via high-performance liquid chromatography and mass spectrometry. Light exposure and berry and air temperature were monitored in Corot noir throughout the season to represent generalized vine microclimate. No significant difference in total anthocyanin concentration was found between exposed and shaded berries, although variable differences in individual anthocyanins were observed by cultivar.
Conclusion and significance This is the first reported characterization of Corot noir berry anthocyanins, and the first assessment of the effect of leaf pulling on anthocyanin concentrations in Corot noir, Maréchal Foch, and Marquette. Like similar studies of V. vinifera grapes, it suggests the lack of a universal correlation between fruit exposure and the quality and quantity of berry anthocyanins.
Introduction
The development of interspecific hybrid cultivars designed to withstand environmental challenges has made commercial winegrape growing viable in regions of the world previously considered ill-suited for such production (Sabbatini et al. 2013, Atucha et al. 2018). Corot noir, Maréchal Foch, and Marquette are three commonly planted cultivars from three major categories of interspecific hybrid grapes: neo-American, French-American/direct producer, and Vitis riparia-based cold hardy, respectively. Nursery catalogs suggest that Corot noir is particularly relevant in the northeastern United States, including Pennsylvania and New York, as well as in Ohio, Indiana, and Illinois (https://www.fulkersonwinery.com/fruit/corot-noir/, https://www.amberggrapevines.com/varieties/corot-noir). Maréchal Foch and Marquette are commonly grown in colder states such as Vermont, midwestern states such as Minnesota, and in Canada (Atucha et al. 2018). These cultivars are capable of producing high quality “vinifera-like” wines (Myles 2013, Pedneault and Provost 2016, Bradshaw et al. 2018, Egorov et al. 2020).
Anthocyanins are water-soluble molecules that contribute red, purple, and blue colors to red grapes and wine (Ribéreau-Gayon et al. 2006). Color is a critical parameter of wine quality determination; whether consciously or unconsciously, the consumer first analyzes wine by its color (Cheynier et al. 2006, Moreno-Arribas 2009, Lawless and Heymann 2010, Aleixandre-Tudo et al. 2015). Interspecific hybrids and native American cultivars can produce sizeable proportions of both mono- and diglucoside anthocyanins, while Vitis vinifera produce only monoglucosides (He et al. 2012, Waterhouse et al. 2016). Hybrid wines may have a purple or blue hue rather than the traditional red or brick, likely due to lower proportions of malvidin anthocyanins (Romero et al. 2008, Manns et al. 2013). A further understanding of the color composition of, and reactions between, components in interspecific hybrid wines will aid producers in optimizing their color expression.
Application of different viticultural techniques can alter the concentration of anthocyanins and other phenolics in grapes, juice, and wine (Downey et al. 2006, Hickey and Wolf 2019, Del-Castillo-Alonso et al. 2020, Gutierrez-Gamboa et al. 2021). In general, less sun exposure results in lower levels of anthocyanins and other flavonoids (Papouskova et al. 2011, Song et al. 2015, Del-Castillo-Alonso et al. 2020). Higher temperature can increase the anthocyanin production to a point, but at ~30°C, many metabolic processes in grapevines stop or slow significantly (Mori et al. 2007, Jones 2014, Shinomiya et al. 2015, Torres et al. 2020, Gutierrez-Gamboa et al. 2021, VanderWeide et al. 2022). In the limited body of work separating the effects of light and temperature, changes in temperature have been shown to influence anthocyanin biosynthesis, as a range of berry temperature exists for optimal metabolic processing and higher berry temperatures can limit anthocyanin accumulation (Bergqvist et al. 2001, Spayd et al. 2002, Downey et al. 2006, Mori et al. 2007, Tarara et al. 2008). However, optimal ranges of light and temperature for biosynthesis vary by cultivar (Morris et al. 2004, Downey et al. 2006, Scharfetter et al. 2019). Temperature differences between exposed and shaded berries ranging from 0 to 13°C have been observed, with larger differences in warmer regions (Spayd et al. 2002, Zhuang et al. 2014, Hickey and Wolf 2019). Shifts in anthocyanin composition can occur based on temperature and light, with the potential for increased biosynthesis of trihydroxylated anthocyanins (delphinidin, petunidin, and malvidin) over that of dihydroxylated anthocyanins (cyanidin and peonidin), and a possible increase in acylated anthocyanins in shaded samples (Chorti et al. 2010, Matsuyama et al. 2014, VanderWeide et al. 2020, Yan et al. 2020, Gil-Munoz et al. 2021, Duan et al. 2022). However, the effect of viticultural practices on anthocyanin accumulation has been researched almost exclusively with V. vinifera cultivars, and many of the studies on berry temperature effects occurred in warm climate regions. Some viticultural treatments have been tested on hybrid grapes but results have been variable; while there is evidence that anthocyanin accumulation in interspecific hybrids can be affected by viticultural practices, the discrete effects of light and temperature are poorly understood (Morris et al. 2004, Sun et al. 2011, 2012, Balint and Reynolds 2017, Scharfetter et al. 2019).
This research aims to characterize the composition of skin anthocyanins in three interspecific cool-climate hybrid grape cultivars (Vitis spp.): Corot noir, Maréchal Foch, and Marquette. Light exposure and berry and air temperature were monitored in Corot noir in 2018 and 2019 to represent generalized vine microclimate. Monomeric skin anthocyanins were extracted from shaded and unshaded fruit from each cultivar at multiple times between veraison and harvest and analyzed via high-performance liquid chromatography (HPLC), with identities later verified through mass spectrometry (MS), to explore the evolution of anthocyanins over the course of ripening and to clarify the effects of vine microclimate on berry color.
Materials and Methods
Viticultural data
In 2018 and 2019, grape samples were collected from three vineyards located in the Finger Lakes region of New York: Corot noir from the Finger Lakes Teaching and Demonstration Vineyard (Penn Yan, NY; 42°42′N; 76°58′W), Maréchal Foch from Prejean Vineyard (Penn Yan, NY; 42°40′N; 76°57′W), and Marquette from Stever Hill Vineyard (Branchport, NY; 42°35′N; 77°9′W). All vineyards are in Yates County within the Finger Lakes American Viticultural Area and are classified as cool-climate, with growing degree days (GDDs) equivalent to Winkler Classification II. Vines at each location were trained on a high wire trellis system and cane-pruned. Further viticultural data and pertinent weather data are provided (Tables 1 and 2, respectively). A randomized complete block design with paired treatment plots was used to accommodate a limited number of data loggers for this trial. At each vineyard, three replicated blocks of treated (exposed [E]) and control (shaded [S]) vines were assigned. E and S treatments were set up as adjacent panel pairs and the locations of these paired panels were randomized within each vineyard. Each experimental unit consisted of a single panel containing four vines. For the E treatments, four to five basal leaves per bearing shoot were removed by hand on both sides of the canopy following fruit set; no leaf removal was performed in the S treatment. Photosynthetically active radiation (PAR) and berry and ambient air temperature were measured in the Corot noir vineyards for 2018 and 2019 to represent generalized vine microclimate. Measurements were taken with an SQ-421 PAR Digital Sensor mounted on a polyvinyl chloride (PVC) tube attached horizontally on the top trellis wire within the canopy and recorded with a Model CR1000X data logger (Apogee Instruments; Campbell Scientific). Berry temperature was measured using six ST-200 fine wire thermistor probes (two probes per replicate) which were placed in one berry on each side of the canopy (Apogee Instruments). Values for PAR and berry temperature were measured every 5 min then averaged and recorded every 15 min. Air temperature within the canopy was recorded every 30 min using Thermochron iButton data loggers (iButtonLink) placed in PVC shelters hung on the top wire of the trellis at a height of ~1.8 m.
Viticultural data for Corot noir, Maréchal Foch, and Marquette vineyards.
Growing degree days (GDDs) and rainfall in Corot noir, Maréchal Foch, and Marquette vineyards during the 1 April to 31 Oct season in 2018 and 2019 (http://newa.cornell.edu/).
Sample collection
Beginning 2 wk after veraison, one composite sample of 100 randomly-sampled berries was collected from each of the six blocks at each of the three vineyards (T1). Sampling was repeated every 2 wk until harvest at times two, three, and four (T2, T3, T4) between veraison and harvest each year, as dictated by season length (Table 3). Berries were blind-picked by hand from different parts of the clusters (tail, shoulder, etc.) and the vine (interior and exterior clusters) within the designated block, bagged, transported in a cooler with ice packs, and frozen at −20°C until needed.
Dates of veraison (50%), sampling time (T1 to T4), and harvest for three grape cultivars in 2018 and 2019. A dash indicates grapes that were no longer available for sampling at a given time due to prior harvesting.
Solvents
All solvents, including hydrochloric acid, methanol, Type 1 water, phosphoric acid, ethyl acetate, and acetonitrile (Fisher Chemical or VWR International, Inc.) were HPLC grade.
Pulp and skin sample preparation
Samples were removed from the freezer (−20°C) and 100 berries were counted and weighed in a tared plastic cup on a balance to record average berry weight (Supplemental Table 1), then were returned to the freezer until processing. To separate skins and pulp, samples were placed in the refrigerator (2°C) to gradually thaw, then berries were individually squeezed between thumb and fore-finger to eject the pulp. Skins were stored in labeled 50-mL foil-wrapped Falcon tubes and returned to the freezer. Pulp was put into a stomacher bag and placed in a Seward Stomacher 400 Circulator for 2 min at 300 rpm, tested for total soluble solids (as Brix) with a Misco Palm Abbe #PA201 Digital Refractometer (Supplemental Table 1), and transferred to labeled clear plastic bags, which were stored in an aluminum envelope in the freezer.
Aluminum weigh boats were labeled and folded to fit six within 45 × 18.5 cm freeze-dryer trays. Skins were taken directly from the freezer and spread in a single layer on an aluminum boat, then run in a 25-hr freeze-dryer cycle in a Medium Home Freeze Dryer (Harvest Right) in 2018 and in a Thermocouple Vacuum Gauge Freeze Dryer (The Virtis Company) in 2019. Once dry, skins were put into labeled clear plastic bags. Each sample was weighed and stored in an aluminum envelope in the freezer (Supplemental Table 1). Samples were protected from light exposure with low ambient lighting and aluminum foil covering throughout processing. Freeze-dried skins were ground in a Retsch MM200 MixerMill Ball Mill (Verder Scientific, Inc.) for 110 sec at 25 rev/sec, then transferred into labeled 15-mL plastic Falcon tubes. Grinder capsules and balls were cleaned with ethanol and thoroughly dried between samples.
Methanol extraction
A modified extraction method was used based on previous literature (Jeffery et al. 2008, Manns and Mansfield 2012). Freeze-dried ground skin samples were removed from the freezer and 3 g were added to a foil-wrapped 250-mL Erlenmeyer flask with 50 mL of 80% methanol solution; after mixing, it was sealed with a neoprene stopper and secured in an ice bath in a Branson 2800 sonicator. Samples were sonicated for 20 min and swirled at 5-min intervals to prevent cavitation and standing waves. After sonication, the sample was filtered through a Buchner funnel with a #1 filter into a foil-wrapped collection flask. The filter cake was washed with a 25-mL aliquot of 80% methanol and returned to the Erlenmeyer flask with 50-mL 80% methanol, then sonication, filtering, and rinsing were repeated. All filtrate in the collection flask was transferred to a round-bottom flask, using 25-mL 80% methanol to rinse the previous collection flask. Filtrate was reduced to 10 to 30 mL using a Büchi Rotavapor R-200 rotary evaporator at 25 kPa pressure and moderately fast speed in a 32°C water bath. The filtrate was then diluted to 50 mL with Type I water and kept frozen in a foil-wrapped Falcon tube until further analysis.
Solid-phase extraction
Samples that had completed methanol extractions were removed from the freezer, defrosted, shaken by hand, and sonicated for 5 min. Two solid-phase extraction (SPE) manifolds, J.T. Baker, Inc. (Avantor) and Burdick & Jackson (Honeywell), were set up with tube racks and eight sets of Varian Bond Elut solvent reservoirs, funnel adaptors, Oasis HLB 3 cc 60-mg cartridges (Waters), and stopcocks. A previously described SPE protocol (Manns and Mansfield 2012) was used to collect the F1A anthocyanin fraction, modified such that samples were placed in a test tube rack in a 30°C water bath under a continuous stream of nitrogen (15 to 30 PSI) during drying.
HPLC-diode array detector analysis
HPLC conditions and the method previously described (Manns and Mansfield 2012) were followed for anthocyanins, full range (pentafluorophenyl [PFP]) using an Agilent 1260 Infinity series HPLC (Agilent Technologies) consisting of a G1322A inline degassing unit, a G1312B binary pump, a G1329B autosampler, a G1316A thermostat column compartment, and a G4212B diode array detector fitted with a 10-mm path, 1-μL volume Max-Light cartridge flow cell. Samples were analyzed on a Kinetex Core-Shell 100 × 2.1 mm PFP column packed with 2.6-mm diameter particles with a 100 Å pore size fitted with an inline Krudkatcher guard filter (Phenomenex). Standards of oenin chloride (malvidin-3-O-glucoside) and malvin chloride (malvidin-3,5-glucoside) were purchased from Extrasynthese. Dilutions of 5, 10, 50, 100, 500, and 1000 mg/L were prepared using 0.01 N hydrochloric acid for quantification. Standards were filtered through a 0.2-μm polyethersulfone (PES) filter and analyzed using the same HPLC protocol as the samples.
Liquid chromatography-mass spectrometry analysis
A subset of 12 samples from each cultivar and timepoint were submitted to the Bioinformatics Research Center at Cornell University for MS analysis. This analysis was performed using an Exion LC (Sciex) equipped with a Kinetex Core-Shell 100 × 2.1 mm PFP column packed with 2.6-mm diameter particles with a 100 Å pore size fitted with an inline Krudkatcher guard filter (Phenomenex) and a Sciex X500B MS operated in ESI positive ion FT mode. Five anthocyanin standards (malvidin-3-O-glucoside, malvidin-3,5-diglucoside, cyanidin-3-O-glucoside, petunidin-3-O-glucoside, and pelargonidin-3-O-glucoside; Extrasynthese) were used to optimize the liquid chromatography-tandem mass spectrometry (LC-MS/MS) method and a calibration curve was created for each. Compound identification for each of these five standards, as well as for other anthocyanin monoglucosides, diglucosides, and acetyl and coumaroyl adducts, were compared to the HPLC data based on mass.
Statistical analyses
Analysis of variance (ANOVA) was performed for each cultivar using RStudio software ver. 1.2.5042 (RStudio Team 2020) to determine significant differences at p < 0.05, <0.01, and <0.001 for anthocyanin concentration among sample time, treatment, and the interaction of time and treatment. A linear mixed effect model using the Kenward-Roger approach to estimate denominator degrees of freedom was used, with fixed effects of time, treatments, and the interaction, and a random effect of panel to control for the repeated measures. The mixed model was fit to the log-transformed quantification values to ensure that the model assumptions of normality and homogeneous variance were met.
Results
Growing seasons
The Corot noir and Maréchal Foch were grown within 9 km of each other in vineyards with similar row orientation (north to south) and aspect (mild east-facing slopes) on the west side of Seneca Lake; growers report comparable climatic conditions most years. The Marquette site, located on Keuka Lake, has an east-west row orientation with a mild east-facing slope and is historically a cooler site than the other two sites. In 2018, GDD accumulation at all sites was within 5% of their long-term averages (Table 2). Rainfall at the Marquette site was ~17% below normal that season, while rainfall at the Corot noir and Marquette vineyards were 4% below average and 12% above normal, respectively. GDD accumulation in 2019 was ~15% below average at all three sites, while rainfall totals were all within 5% of the 10-yr averages.
Viticulture
PAR in the Corot noir E vines was 14.5 times higher in 2018 (E maximum = 1262.8; E average = 139.1; S maximum = 269.7; S average = 9.6) and nine times higher in 2019 (E maximum = 1273.3; E average = 120.1; S maximum = 748.1; S average = 13.2) than the corresponding PAR in the S vines for the period between 2 wk postveraison to harvest. Average temperatures of E and S berries varied significantly both years, with greater variance from higher temperature in E berries in 2019 (Table 4).
Viticultural effects of Corot noir treated (exposed [E]) and control (shaded [S]) samples in 2018 and 2019.
Total skin anthocyanin content
There were no significant differences in total anthocyanins between E and S samples for any cultivar or year, with the exception of Maréchal Foch, which in 2019 had a greater concentration of total anthocyanins in E samples (p = 0.0281) (Tables 5 to 7). As reported below, the concentration of some individual anthocyanins in E samples varied by cultivar and vintage.
Skin anthocyanin profiles in Corot noir treated (exposed [E]) and control (shaded [S]) samples in 2018 and 2019. DW, dry weight; glu, glucoside; diglu, diglucoside.
Skin anthocyanin profiles in Maréchal Foch treated (exposed [E]) and control (shaded [S]) samples in 2018 and 2019. DW, dry weight; nd1, not detected in any samples; nd2, not detected in samples at final timepoint before harvest; glu, glucoside; diglu, diglucoside.
Skin anthocyanin profiles in Marquette treated (exposed [E]) and control (shaded [S]) samples in 2018 and 2019. DW, dry weight; glu, glucoside; diglu, diglucoside.
Skin anthocyanin profiles
Ten peaks which appeared both years in HPLC and MS data were labeled Peak 1 (P1) through 10 (P10) (Figure 1) and were tentatively identified through comparison to standards, MS data, and reported elution order (Liang et al. 2008, He et al. 2012, Manns and Mansfield 2012). In all samples, the peak group following the elution of malvidin-3-O-glucoside was identified as “modified anthocyanins” (MA), previously defined as various acetylated and coumaroylated anthocyanin mono- and diglucosides and presented as a group (Manns and Mansfield 2012). Tentative anthocyanin identifications (TIDs) in Corot noir, Maréchal Foch, and Marquette are shown (Tables 5, 6, and 7, respectively). Quantification was calculated via average peak area and is represented in mg/g dry grape skin of malvidin-3-O-glucoside equivalents. P1, P2, P3, P5, and P6 are labeled as ‘Unknown A’ through ‘Unknown E’ because the identity of each could be one of two anthocyanins; Unknown A is either cyanidin-3,5-diglucoside or delphinidin-3,5-diglucoside; Unknowns B and C are either delphinidin-3-O-glucoside or petunidin-3,5-diglucoside; and Unknowns D and E are either peonidin-3,5-diglucoside or pelargonidin-3-O-glucoside.
Representative high-performance liquid chromatography chromatograms at 520 nm of skin anthocyanin extract from three grape cultivars.
Corot noir
The highest anthocyanin concentration in Corot noir was the MA group, followed by P4 (TID cyanidin-3-O-glucoside), then P3, P7, and P8 (TIDs Unknown C; malvidin-3,5-diglucoside; and petunidin-3-O-glucoside; respectively) (Figure 2). In 2018, exposed Corot noir samples had higher concentrations of P1 (p = 0.0077), P2 (p = 0.0105), and P3 (p = 0.0266) (Unknowns A, B, and C, respectively), and lower concentrations of P8 (p = 0.0117) and P10 (p = 0.0028) (TIDs petunidin-3-O-glucoside and malvidin-3-O-glucoside, respectively). In addition, there was an interaction of timepoint and treatment (p = 0.0218) for MA with a difference between E and S at T4 (p = 0.0124). In 2019, Peak 10 (p = 0.0281, TID malvidin-3-O-glucoside) was lower in E samples, and MA again showed an interaction of timepoint and treatment (p = 0.0109), with a difference between E and S at T4 (p = 0.0004).
Quantification of the four most dominant individual anthocyanins and the group of modified anthocyanins during ripening in Corot noir treated (exposed [E]) and control (shaded [S]) samples in 2018 and 2019. Error bars signify standard deviation across triplicate samples. DW, dry weight.
Maréchal Foch
In both vintages of Maréchal Foch, the highest anthocyanin concentrations were P4, P8, P10 (TIDs cyanidin-3-O-glucoside; petunidin-3-O-glucoside; and malvidin-3-O-glucoside; respectively), and the MA group (Figure 3). Peaks 1 to 10 did not appear in every Maréchal Foch sample; for example, in 2018, P1 and P2 did not appear at T1 or T3, and P5 and P6 did not appear at T1. Similarly, in 2019 P1 was not observed at T2 or T3, and P2 was completely absent. In 2018, concentrations of P9 (p = 0.0295, TID peonidin-3-O-glucoside) and P10 (p = 0.0266, TID malvidin-3-O-glucoside) were lower in E (Figure 3). In 2019, E samples showed higher accumulation of P3 (p = 0.0004), P4 (p = 0.0198), P5 (p = 0.0056), and P7 (p = 0.0007) (TIDs Unknown C; cyanidin-3-O-glucoside; Unknown D; and malvidin-3,5-diglucoside; respectively).
Quantification of the three most dominant individual anthocyanins and the group of modified anthocyanins during ripening in Maréchal Foch treated (exposed [E]) and control (shaded [S]) samples in 2018 and 2019. Error bars signify standard deviation across triplicate samples. DW, dry weight.
Marquette
Within the Marquette anthocyanin profile, Peaks 4, 7, 8, 10 (TIDs cyanidin-3-O-glucoside; malvidin-3,5-diglucoside; petunidin-3-O-glucoside; and malvidin-3-O-glucoside; respectively) and MA had the highest concentration, although these proportions changed over the two vintages (Figure 4). In 2019, Marquette samples were collected at two timepoints (rather than three) because of an unexpectedly early harvest of the commercial vineyard block. Exposure treatment made no difference in the accumulation of individual anthocyanins in either year, but in 2019, Peaks 7, 8, and 10 had a timepoint × treatment interaction (p = 0.0438, 0.0225, and 0.0385, respectively), with accumulation lower in E samples at T1 (p = 0.0468, 0.0194, and 0.0142, respectively).
Quantification of the four most dominant individual anthocyanins and the group of modified anthocyanins during ripening in Marquette treated (exposed [E]) and control (shaded [S]) samples in 2018 and 2019. Error bars signify standard deviation across triplicate samples. DW, dry weight.
Discussion
Viticultural effects
Traditionally, selected leaves are removed along shoots to increase fruit exposure to radiation and heat over the maturation process (VanderWeide et al. 2022). In this study, PAR was nine to 14.5 times higher in Corot noir E samples in both years, concurrent to a difference in berry temperature (Table 4). Although light is essential for anthocyanin biosynthesis, temperature variation seems more influential in color difference than light variation, with optimal color accumulation occurring at 20 to 30°C (Haselgrove et al. 2000, Bergqvist et al. 2001, Spayd et al. 2002, Mori et al. 2007, Downey et al. 2008, Tarara et al. 2008, Jones 2014, Shinomiya et al. 2015, VanderWeide et al. 2020). It is challenging to separate the effects of light and heat because increased PAR in warm region studies is rarely decoupled from berry warming. Prior studies reporting increased anthocyanin content with leaf removal occurred in warm areas such as southern Spain, California, or Mediterranean wine regions (Poni et al. 2006, Tardaguila et al. 2010, Diago et al. 2012a, 2012b, Gatti et al. 2012, Kotseridis et al. 2012, Intrigliolo et al. 2014, Cook et al. 2015, Chorti et al. 2016, Pavic et al. 2019, Anić et al. 2021, Moreno et al. 2021, Mucalo et al. 2021, Stefanovic et al. 2021), where light exposure also increased berry temperature. The best grape ripening conditions occur when temperatures are high enough to reach optimal maturity (25 to 30°C), but not so high that metabolic processes slow (>30°C) (Mori et al. 2007, Tarara et al. 2008, Azuma et al. 2012, VanderWeide et al. 2020, Gutierrez-Gamboa et al. 2021). In V. vinifera studies where berry temperature was increased independent of PAR through direct heating, incident radiation, or higher air temperature, anthocyanin accumulation was lower at temperatures exceeding 27 to 30°C, while berry heating below 27°C coupled with light exposure showed sufficient anthocyanin accumulation (Haselgrove et al. 2000, Bergqvist et al. 2001, Spayd et al. 2002, Downey et al. 2006, Shinomiya et al. 2015, Song et al. 2015, Del-Castillo-Alonso et al. 2020).
In the current study, average berry temperatures in Corot noir stayed below 20°C, and therefore below the low end of optimal color accumulation range (Table 4), suggesting that the PAR increase from leaf removal was inadequate to produce a significant difference in berry temperature in the cool climate of the Finger Lakes region. It must be noted that temperature was only measured in exterior berries, while 100-berry samples were selected throughout the cluster, so reported berry temperature is somewhat higher than the average temperature of all fruit on the cluster. If the berries with the greatest exposure averaged less than 20°C, it can be assumed that more shaded berries in the same cluster would be even colder. Wind velocity, which can cool the grapes despite incident radiation, may have played a role, but was not recorded (Spayd et al. 2002). In addition, the effects of both light and temperature on anthocyanin accumulation are critically cultivar dependent (Downey et al. 2006, Lo Cicero et al. 2016) and have not been independently examined in non-vinifera cultivars. Furthermore, measurements in one vineyard might not be representative of others, as even slightly different environmental factors can affect microclimatic characteristics (Reynolds 2010). It is important to highlight that previous studies have focused on V. vinifera cultivars trained using vertical shoot-positioning (VSP), while all three cultivars in this study were trained using a high wire system. As a result, the fruiting zone of these vines is more exposed to sunlight in the hours immediately around solar noon because fruiting buds are at the top of the trellis and exposed to sunlight from directly above. Thus, the timing and possibly the amount of solar radiation acting on fruit varies by trellis type. This suggests that climate and cultivar-specific data on anthocyanin accumulation cannot be generalized across regions and species and further work is needed to differentiate light and temperature effects on anthocyanin accumulation in cool-climate, interspecific hybrid grapes.
Skin anthocyanin profiles
Total anthocyanins
The lack of difference between E and S total anthocyanin concentrations in Corot noir and Marquette for both years, and in Maréchal Foch in 2018, is comparable with several reports of leaf removal treatments in interspecific hybrids (Le et al. 2022) and V. vinifera cultivars (Matus et al. 2009, Chorti et al. 2010, Petropoulos et al. 2011, Pisciotta et al. 2013, Baiano et al. 2015, Sivilotti et al. 2016, Pastore et al. 2017, Sun et al. 2017, Carew et al. 2020, Torres et al. 2020, Le et al. 2022, Tessarin et al. 2022). A Merlot (V. vinifera L.) leaf removal study showed no effect on anthocyanin accumulation following severe defoliation, but variable effects followed a moderate treatment, with anthocyanin concentration significantly higher in cool-climate grapes but somewhat lower in the warmer region (VanderWeide et al. 2022). The authors suggested that factors other than PAR and berry temperature, such as GDDs between veraison to harvest, may have a greater effect than defoliation on anthocyanin accumulation. Several studies suggest that the manner of defoliation is relevant, with mechanical leaf removal showing greater impact on anthocyanin accumulation than manual leaf removal (Diago et al. 2012a, VanderWeide et al. 2018, 2020). The increase in total anthocyanins in E samples for Maréchal Foch in 2019 demonstrates the complex effect that vintage and cultivar can have on the accumulation of anthocyanins. The annual variability in climate, including fewer GDDs, higher maximum ambient temperature, and a wider range in temperature in 2019 (Tables 2 and 4), may have influenced total anthocyanin accumulation in Maréchal Foch in 2019, but it is unclear why it was the only cultivar affected.
Variable effects on anthocyanins have also been reported for other viticultural treatments. In prior work on Maréchal Foch and Corot noir, shoot and cluster thinning did not reliably change wine anthocyanin content (Sun et al. 2011, 2012). In contrast, an analysis of juice and wine from three V. riparia-based cold-climate cultivars (Frontenac, Marquette, and Petite Pearl) found that total phenolic content and monomeric anthocyanin concentration increased in both juice and wine with a single preveraison leaf and lateral shoot removal treatment, while cluster and shoot thinning in French-American hybrids Chancellor and Villard noir resulted in greater red color (Morris et al. 2004, Scharfetter et al. 2019). Similarly, larger increases in wine anthocyanins were observed in Merlot, Cabernet Franc, and Cabernet Sauvignon following cluster thinning in conjunction with leaf removal, compared to leaf removal alone (Di Profio et al. 2011). In short, evidence suggests that cultivar, vintage, defoliation timing, and microclimate all influence the efficacy of leaf removal as a means to enhance anthocyanin accumulation (Lee and Skinkis 2013, Pastore et al. 2013, Sivilotti et al. 2016, Ćirković et al. 2019, Pavic et al. 2019, Ivanišević et al. 2020, Mucalo et al. 2021, Tessarin et al. 2022).
Individual anthocyanins
The similarities in total anthocyanin concentrations of E and S treatments belies differences in individual anthocyanins, which were evident in both Corot noir and Maréchal Foch (Tables 5 and 6). Previous work has shown that distribution of individual anthocyanins can vary greatly by viticultural treatment without affecting total wine anthocyanin concentrations, with low predictability as to which types of anthocyanins respond to treatment (dihydroxylated or trihydroxylated, monoglucosides or diglucosides, acylated or non-acylated), if a strict trend is observed at all (Guidoni et al. 2008, Tarara et al. 2008, Rustioni et al. 2011, King et al. 2012, Matsuyama et al. 2014, Baiano et al. 2015, Feng et al. 2015, Osrečak et al. 2016, Pavic et al. 2019, Ivanišević et al. 2020, Stefanovic et al. 2021, Yue et al. 2021, Duan et al. 2022, Tessarin et al. 2022). Berry exposure to solar radiation has been found to affect the biosynthesis of tri- and di-substituted anthocyanins and their ratio (Torres et al. 2020, VanderWeide et al. 2020), with E samples showing higher concentrations of dihydroxylated anthocyanins (cyanidin and peonidin) and lower concentrations of trihydroxylated (delphinidin, petunidin, and malvidin) and acylated anthocyanins (Chorti et al. 2010, Matsuyama et al. 2014, VanderWeide et al. 2020, Yan et al. 2020, Gil-Munoz et al. 2021, Duan et al. 2022).
In the current work, dihydroxylated anthocyanins cyanidin-3-O-glucoside and Unknown D were higher in exposed Maréchal Foch in 2019, and some E samples had lower concentrations of trihydroxylated anthocyanins petunidin-3-O-glucoside (Corot noir 2018) and malvidin-3-O-glucoside (Corot noir 2018, 2019; Maréchal Foch 2018) (Figures 2 and 3). Exposed samples of Corot noir also had lower MA at T4 in both years. In contrast to prior work, however, dihydroxylated peonidin-3-O-glucoside was lower in E samples of both Corot noir and Maréchal Foch in 2018, and E samples had higher concentrations of trihydroxylated Unknowns B and C in Corot noir in 2018, and of Unknown C and malvidin-3,5-diglucoside in Maréchal Foch in 2019. The concentration of Unknown A, which is either cyanidin-3,5-diglucoside (dihydroxylated) or delphinidin-3,5-diglucoside (trihydroxylated), was also higher in Corot noir in 2018. It is unclear why some individual concentrations varied from prior results, although cultivar and climate differences may be at play. The differing results between the two vintages of the same cultivar suggest the effect of season and the potential inconsistency of viticultural practice effects on anthocyanin biosynthesis changes, as reported previously (Reynolds et al. 2006, Frioni et al. 2017, Blancquaert et al. 2019, Drenjančević et al. 2023).
Potential wine color
Because of unique anthocyanin profiles in the berry, wine color in Corot noir, Marquette, and Maréchal Foch may vary from that of traditional V. vinifera wine, so elucidating anthocyanin composition can help predict eventual color expression. The large proportion of MAs in all three cultivars (Tables 5 to 7) could result in more stable wine color, as acylated anthocyanins are more resistant to chemical modification than non-acylated (Chen et al. 2022). Because cyanidin-3-O-glucoside was the most prevalent anthocyanin in Corot noir, the resulting wine color would likely vary from that of V. vinifera, where malvidin-3-O-glucoside dominates (Romero et al. 2008, He et al. 2012, Manns et al. 2013, Jackson 2016). Cyanidin contributes a pinker hue than the more purple malvidin, so Corot noir wine is likely to be lighter in color than V. vinifera (Boulton 2001, He et al. 2012, Ananga et al. 2013, Manns et al. 2013, Burtch et al. 2017). High concentrations of cyanidin-3-O-glucoside and petunidin-3-O-glucoside in both Maréchal Foch and Marquette could similarly result in lighter-colored wines. V. vinifera grapes are unique in producing virtually none of the anthocyanin diglucosides characteristic of interspecific hybrid fruit; diglucosides are less pigmented than monoglucosides at wine pH (Boulton 2001, Burtch et al. 2017), so the proportion of mono- to diglucosides can influence wine color intensity. Higher concentrations of malvidin-3,5-diglucoside in Corot noir and Marquette suggest wines with less intensity than Maréchal Foch; as found in prior work, Maréchal Foch contains a relatively low concentration (Kontić et al. 2016). Finally, the unique attributes of these color profiles could also have an effect on color stability reactions throughout wine aging, as diglucosides are less reactive than monoglucosidic anthocyanins (Burtch et al. 2017). However, studies have shown that changes in anthocyanin composition in grapes at harvest do not always demonstrate corresponding changes in the anthocyanin composition of the final wine (King et al. 2012, Chorti et al. 2016, Drenjančević et al. 2023). The lack of significant difference in total anthocyanins in E and S berries suggests that berry exposure would not result in perceptible modification of wine color in the cultivars studied, but winemaking trials are required to assess potential variation due to changes in individual anthocyanin concentration.
Conclusion
E and S berries of three interspecific hybrid winegrape cultivars showed some difference in individual skin anthocyanin concentrations, but differences generally were not seen in total concentration. This may be due in part to the lack of difference in exposed berry temperature despite a significantly higher level of PAR. Annual differences in anthocyanin concentrations suggest that other factors like vintage, cultivar, or viticultural practices may have a greater effect on anthocyanin development than berry exposure. Consequently, the cost and labor of leaf removal in cool-climate Corot noir, Maréchal Foch, and Marquette vineyards may be wasted in an attempt to alter anthocyanin accumulation, although it may still be a beneficial practice for other reasons such as disease management. Despite these findings, the possibility of modifying concentrations of specific anthocyanins is worth additional exploration. Furthermore, the difficulties of separating the effects of PAR and heat on berry development and of generalizing viticultural results across climate and grape species is evident. This work adds to the limited knowledge of anthocyanin evolution in interspecific hybrid cultivars, which can spur further optimization of anthocyanin composition and color in the cool-climate red wines they produce.
CRediT Authorship Contributions
CD: Data Curation; CD, HWP, AKM: Writing – Original Draft, Writing – Review & Editing; CD, AKM: Formal Analysis, Investigation; HWP: Resources; HWP, AKM: Conceptualization, Funding Acquisition, Methodology; AKM: Project Administration
Supplemental Data
The following supplemental materials are available for this article in the Supplemental tab above:
Supplemental Table 1 Average (n = 100) weight of berries and dried berry skin following sample preparation, and total soluble solids (TSS) of berry pulp ± standard deviation, across replicates for all three cultivars at each sampling timepoint and each treatment (exposed [E] and control [shaded; S]) in 2018 and 2019. Marquette was only sampled twice in 2019.
Data Availability
The data underlying this study are available on request from the corresponding author.
Footnotes
2018 NYSAES Research Venture Grant for funding; Finger Lakes Teaching and Demonstration Vineyards, Prejean Vineyards, and Stever Hill Vineyards for sample donation; Don Caldwell for vineyard management; Lynn Marie Johnson for statistics consultation; Ruchika Bhawal at the Cornell Biotechnology Research Center for mass spectrometry analysis.
Dadmun CH, Walter-Peterson HC and Mansfield AK. 2026. Effect of sun exposure on the evolution and distribution of skin anthocyanins in interspecific red hybrid winegrapes. Am J Enol Vitic 77:0770016. DOI: 10.5344/ajev.2026.25046
By downloading and/or receiving this article, you agree to the Disclaimer of Warranties and Liability. If you do not agree to the Disclaimers, do not download and/or accept this article.
- Received September 2025.
- Accepted April 2026.
- Published online July 2026
This is an open access article distributed under the CC BY 4.0 license.
References
Vol 77 Issue 2
Jump to section
Related Articles
Cited By...
More from this TOC section
Similar Articles
