Abstract
Background and goals Early leaf removal (ELR) is a canopy management practice that can reduce bunch rot and improve fruit maturity by modifying cluster architecture and fruit zone microclimate. The goal of this study was to identify an ELR threshold that balances improvements to fruit quality and reductions in yield for a highly vegetative and high-yielding Vitis vinifera cultivar grown in the eastern United States, Grüner Veltliner.
Methods and key findings Over two consecutive years, we applied a wide range of ELR severities (three to 12 leaves per shoot) at trace bloom (Eichhorn-Lorenz [E-L] stage 19) and evaluated relationships between increasing ELR severity and fruit set, yield parameters, fruit maturity, bunch rot, and juice volatile composition at harvest. In the first year, increasing ELR gradually reduced fruit set, subsequently lowering yield via reducing berry number and cluster weights. Additionally, ELR increased fruit ripeness while reducing rot incidence. However, extreme ELR (10 to 12 leaves per shoot) negatively affected inflorescence floret number and fruitfulness in the following year. In the second year, only high ELR severities (eight to 12 leaves per shoot) reduced yield parameters, and improvements in fruit ripeness or bunch rot were not observed, likely because of poor weather conditions during fruit ripening. The impact of increasing ELR severities on juice volatile composition was largely limited to benzaldehyde (a volatile phenylpropanoid), which was negatively affected by ELR in both years.
Conclusions and significance Removal of six to eight leaves per shoot appeared to be the optimal ELR severities for grapegrowers seeking crop management and rot reduction for high-yielding cultivars like Grüner Veltliner. However, as reported in the second year of this study, the effects of ELR may vary by year and be limited by unfavorable weather conditions.
Introduction
Since Grüner Veltliner (Grüner V.; Vitis vinifera L.), a white grape cultivar widely planted in Austria, was first planted in Pennsylvania ~2003, it has received increased interest from the Pennsylvania wine industry (Smith and Centinari 2019). However, only a few studies have explored aspects of Grüner V. grape and wine production in the eastern United States, such as crop management methods (Smith and Centinari 2019) and wine chemical and sensory characteristics (Keller et al. 2021). Grüner V. is a high-yielding cultivar that also tends to exhibit high vegetative vigor under eastern U.S. environmental conditions. Because of this, canopy management practices applied to other white cultivars grown in the region and with lower vigor and crop potential (e.g., Chardonnay, Riesling) may need to be tailored to optimize Grüner V. production and grape quality.
Among the management practices used by grapegrowers in the eastern U.S., fruit zone leaf removal is widely applied to reduce bunch rot and improve fruit quality. Warm and humid conditions, often occurring during the ripening period, favor the development of bunch rot, which can result in significant economic losses if not properly managed. While leaf removal at or after fruit set is still the most popular practice among commercial growers, many studies have explored the use of early leaf removal (ELR), i.e., leaf removal applied just before or at bloom for increased control of bunch rot diseases in several cultivars (reviewed in VanderWeide et al. 2021), including Grüner V. (Smith and Centinari 2019).
Removing a substantial number of basal leaves (five to eight leaves per shoot) at this phenological stage can strongly limit carbohydrate supply to developing inflorescences, which reduces fruit set and alters cluster morphology (i.e., less compact clusters with fewer and/or smaller berries; Poni et al. 2006, Intrieri et al. 2008). This modification of cluster architecture could provide a further benefit for rot control compared to post-fruit set leaf removal, which only improves fruit zone microclimate (Hed et al. 2015). However, some studies in the eastern U.S. have suggested that rot development may be more controlled by improved fruit-zone microclimate than by changes in cluster morphology (Hickey and Wolf 2018, Hed and Centinari 2021). In high-yielding cultivars, ELR can also result in beneficial reductions in fruit mass and concurrent improvement in fruit quality (Gatti et al. 2012); potential undesired yield penalties should be considered, however, especially if not offset by relevant improvements in fruit chemistry and/or reductions in rot (Hed and Centinari 2018). In some studies, ELR also resulted in carryover reductions in yield due to reduced fruitfulness (Sabbatini and Howell 2010, Acimovic et al. 2016, Hickey and Wolf 2018). These reductions in fruitfulness have been attributed in part to carbon limitations during inflorescence primordia development, since ELR is imposed at bloom, approximately when inflorescence primordia initialization begins (Noyce et al. 2016a, 2016b, Lopes et al. 2020).
Previous ELR work on vigorous Grüner V. in the eastern U.S. suggests that heavy levels of leaf removal (>5 basal leaves per shoot) may be necessary to induce a carbohydrate limitation that significantly reduces fruit set and modifies cluster morphology (Smith and Centinari 2019). Moreover, removing five basal leaves at trace bloom in Grüner V. did not induce any recovery mechanisms implicated in ELR-driven improvements in fruit composition, such as increased lateral shoot growth or greater fruit sugar allocation (Poni et al. 2006).
The goal of this study was to identify an optimal ELR rate that balances the goals of rot reduction and fruit quality improvement with potential adverse effects on yield. We used an approach similar to that reported by Acimovic et al. (2016) and imposed six different ELR severities (three to 12 leaves per shoot) on vigorous Grüner V. vines for two consecutive years, to identify relationships between production and fruit quality parameters and ELR severities. To detail the effects of ELR on fruit quality characteristics that can alter wine chemistry and sensory attributes, we analyzed fruit volatile composition. Desirable improvements to grape and wine volatile composition caused by ELR have been widely reported in red cultivars (Vilanova et al. 2012, Verzera et al. 2016, Moreno et al. 2017, Iorio et al. 2022). Comparatively, previous work in white cultivars has only targeted aromatic or semi-aromatic cultivars like Istrian Malvasia, Riesling, and Semillon (Bubola et al. 2009, Komm and Moyer 2015, Alessandrini et al. 2018, respectively). Grape and wine volatile composition and sensory qualities for these cultivars are strongly influenced by concentrations of terpenes (Robinson et al. 2014), a class of higher alcohols sensitive to changes in fruit-zone microclimate (Reynolds and Wardle 1989, Reynolds et al. 1996). However, our objective was to evaluate if ELR can impact juice free volatile composition of a non-aromatic cultivar like Grüner V., as its juice and wine composition is less influenced by a single class of compounds and is instead better characterized by those compounds that are common to most white wines (e.g., higher alcohols, esters, fatty acid derivatives, etc.) (Pavelescu et al. 2012, Keller et al. 2021). These volatiles can still be affected by ELR (Bubola et al. 2009, Alessandrini et al. 2018), largely as a by-product of biochemical responses to microclimatic modification and increased berry sun exposure (Pastore et al. 2013).
Based on previous studies comparing ELR severities in other cultivars and work on ELR conducted on Grüner V. (Acimovic et al. 2016, Smith and Centinari 2019), we hypothesized that ELR severities greater than five leaves per shoot would significantly reduce fruit set, leading to lower cluster weights, lower rot incidence and severity, and advanced fruit technological maturity. However, we also hypothesized that carbon limitation imposed by extreme ELR severities (i.e., 10 to 12 leaves per shoot) would cause an undesirable reduction in yield at harvest, in addition to reductions in yield potential in the following season. While no study has quantified the free volatile composition of Grüner V. juice, we hypothesized that increasingly altered source-sink dynamics and greater sunlight exposure due to higher ELR severity would linearly and positively affect volatile composition, alongside improvements in ripening. This would result in higher concentrations of grape-derived volatiles important for white wine aroma, such as derivatives of alcohols and fatty acids.
Materials and Methods
Vineyard site and experimental design
The study was conducted in 2017 and 2018 on Grüner V. cl. 01 (V. vinifera L.) at a commercial vineyard in Lewisburg, PA (40°996′N; 76°891′W; 171 m asl), planted in 2010. Vines were grafted on to 101-14 Mgt (Vitis riparia × Vitis rupestris) and trained to a bilateral cordon with vertical shoot positioning. Vine spacing was 1.5 m × 2.4 m. In both years, vines were managed according to standard disease and pest management practices for V. vinifera cultivars in the eastern U.S. (Wolf 2008). Vines were spur-pruned to two-bud spurs during winter dormancy in March of each year and were later shoot-thinned to an average of 13 shoots/m in May every year.
Four adjacent rows were selected for the study. The experiment was a randomized complete block design with eight blocks (two blocks per vineyard row) and six ELR treatments. Each block consisted of six panels, each randomly assigned to an ELR treatment. There were four adjacent vines within each panel and data were collected from the two center vines and the inward-facing cordons of the first and last vine of each panel. The same treatments were imposed on the same vines in both 2017 and 2018, to account for assessment of carryover effects of ELR. Leaf removal was implemented in 2017 and 2018 on 12 and 6 June, respectively, when shoots reached an average growth stage of Eichhorn-Lorenz (E-L) 19 (“first flower caps loosening”), according to the modified E-L system (Coombe 1995). On each vine, ELR was applied by manually removing a predetermined number of leaves from the main shoot, beginning at the base of the shoots. All lateral shoots that developed on the same nodes according to the ELR severities were also removed. Six ELR severities were implemented in 2017 and 2018: 0 leaves per shoot (ELR-0), three leaves per shoot (ELR-3), six leaves per shoot (ELR-6), eight leaves per shoot (ELR-8), 10 leaves per shoot (ELR-10), and 12 leaves per shoot (ELR-12). The ELR treatments were applied once, and no follow up leaf removal was performed in either year. Hedging was performed twice during the growing season by the commercial grower.
Seasonal weather conditions
For both years, air temperature and precipitation data were measured at hourly intervals and collected using an onsite weather station accessed via the Network for Environment and Weather Applications (http://newa.cornell.edu). Seasonal heat accumulation was assessed from 1 April to harvest as cumulative daily growing degree days (GDD, base 10°C). Heat accumulation during the ripening period was assessed from 15 Aug to harvest as GDD. The beginning date for GDD calculation was set as 1 April because daily temperatures during April can be high enough for GDD accumulation to occur, even if grapevines are still dormant. The beginning date for GDD calculation for the ripening period was set as 15 Aug because grapes had reached veraison (E-L 35) by this date. A base formula of [GDD = ((Tmax + Tmin)/2) − 10] was used to calculate daily GDD, where Tmax and Tmin represent daily maximum and minimum temperature, respectively. Cumulative rainfall was calculated for these same periods by summing daily rainfall values.
Fruit set and yield parameters
On the day prior to ELR application, three shoots per experimental unit (i.e., one shoot per vine) were flagged for measurement of fruit set and various cluster metrics on basal and distal clusters in 2017 and 2018. The percentage of fruit set was calculated as the ratio between the number of berries counted at harvest to the number of florets (i.e., the individual flowers of the inflorescence) at E-L 19 for basal and distal clusters separately (Poni et al. 2006, Acimovic et al. 2016). Estimation of the actual number of florets at E-L 19 from digital photographs was done using the protocol outlined in Smith and Centinari (2019). Inflorescences were photographed at E-L 19 (on 11 June 2017 and 5 June 2018) and the total number of berries per cluster was counted at harvest for each flagged shoot.
Basal and distal clusters on flagged shoots were harvested separately the day prior to harvesting the remainder of the crop. Clusters were individually bagged and transported to the laboratory where they were weighed and stored at −20°C. Clusters were deconstructed and all berries were counted and weighed to determine the total number of berries per cluster and average berry weight. Rachises were also weighed. Berries were visually inspected and sorted into two categories: healthy, full-size berries and rot-affected berries (all bunch rot variations, such as sour rot and Botrytis cinerea, combined). These numbers were used to determine rot severity, or the % berries per cluster affected by rot. Healthy, full-size berries were used for analysis of total soluble solids (TSS), pH, and titratable acidity (TA), performed according to the instructions described in Smith and Centinari (2019). These measurements were performed on a combined sample of all the berries from basal and distal clusters.
All experimental vines were manually harvested the day prior to commercial harvest on 20 Sept 2017 and 19 Sept 2018. To determine the total yield per vine and average cluster weights, clusters were counted and weighed using a hanging scale with a 0.01 kg accuracy (Pelouze 7710, Rubbermaid, Inc.). Clusters with >5% rot-affected berries were sorted into separate lugs, counted, and weighed separately from healthy fruit. All bunch rot variations were combined to calculate rot incidence (% clusters with rot).
Leaf area measurements
Twice in both years, leaf area measurements were performed on flagged shoots: first, on the day ELR was applied, and later, at harvest. When ELR was applied, the leaf area of all leaves removed from each flagged shoot was measured to estimate the average leaf area removed (cm2 leaf area/shoot) for each ELR treatment. Additionally, to calculate the percentage of leaf area removed with each ELR treatment, 10 non-defoliated shoots from non-experimental Grüner V. vines were sampled from the same rows. Leaves from flagged shoots were bagged and placed on ice, while whole non-experimental shoots were wrapped at their basal end with a moist paper towel and bagged in a large plastic bag. Samples were transported to the laboratory, where leaf area was quantified per main and lateral shoots separately using a LI-3100c scanning leaf area meter (LI-COR Bioscience). The percentage of leaf area removed was then calculated as the ratio between the total leaf area removed per shoot of ELR-treated vines to the total leaf area per shoot of non-experimental vines. On the same day of basal and distal cluster harvest, all three flagged shoots were harvested for end-of-season leaf area measurements, following the same instructions outlined above.
Fruit sugar allocation parameters
Using individual shoot yield data from flagged shoots, fruit sugar allocation was measured by calculating the total sugar allocated to an individual berry, shoot, and cm2 leaf area (Poni et al. 2006). The total sugar per berry was calculated by multiplying the TSS and average berry weight of basal and distal clusters, while the total sugar per shoot was calculated by multiplying the total sugar per berry by the total yield per shoot. To determine the total sugar per cm2 leaf area, the total sugar per shoot was divided by the total leaf area per shoot at harvest.
Juice free volatile composition
On the day prior to harvest in both years, samples of 400 berries were randomly harvested per experimental unit for juice free volatile composition analysis. To avoid contributing rot-derived compounds to juice free volatile composition (Steel et al. 2020), only berries without visible rot or infection were sampled. To work efficiently and effectively within time limitations for flash-freezing berries, a reduced number of samples were evaluated, including ELR-0, ELR-6, ELR-8, ELR-10, and ELR-12 in blocks 1 through 6. Berries were bagged, placed on dry ice, and transported to the laboratory where they were flash frozen with liquid nitrogen and stored at −80°C until volatile composition analysis could be performed. Headspace solid-phase microextraction-gas chromatography-mass spectrometry (HS SPME-GC-MS) was used to assess juice free volatile composition, per the method outlined in Keller (2020). Chromatograms were aligned and processed using PARADISe v. 3.4 (Johnsen et al. 2017). Compounds were tentatively identified via the National Institute of Standards and Technology Mass Spectral Library (version 14) and by referencing calculated Kovats retention indices with previously published values for all compounds. Compound identifications were confirmed using authentic chemical standards for all compounds, except for 3-Ethyl-4-methyl-1-pentanol, due to constraints on standard procurement. Compounds were reported in internal standard (IS) equivalents (μg/L). Volatile compounds were also analyzed by chemical class (Alessandrini et al. 2018) according to classes provided in Robinson et al. (2014) and Waterhouse et al. (2016) based on chemical structure; this was done to characterize the responses of chemical composition and individual compounds to ELR at a broad level.
Assessment of carryover effects of ELR severity
The effects of ELR severity on the percent of live buds and fruitfulness (i.e., the number of clusters per shoot) were assessed in the year following application. Two cordons per experimental unit were randomly selected and the number of buds with emerging shoots and dead buds were counted separately to determine the percentage of live buds. The total number of clusters and shoots were counted for each cordon to determine the number of clusters per shoot. Measurements were conducted on 23 May 2018; it was not possible to repeat this assessment in spring 2019 because a severe winter freeze event (i.e., four consecutive days with air minimum temperatures of −18.3, −19.3, −17.1, and −17.0°C) at the end of January and beginning of February caused widespread bud mortality. On 13 March 2019, two eight-node canes were randomly collected for each experimental unit and transported back to the laboratory to assess bud mortality of primary, secondary, and tertiary buds. All buds per cane were manually dissected using a razor blade and were visually inspected to determine if necrosis had occurred (Wolf 2008).
Statistical analysis
Data were analyzed using SAS v. 9.4 and JMP Pro 16 (SAS Institute, Inc.). To test how increasing ELR severity affected viticultural variables between seasons, mixed model analysis of covariance (ANCOVA) was performed on all viticultural data using PROC MIXED in SAS, according to the multistep processes outlined by Marini et al. (2002) and Marini and Ward (2012). In this analysis, year was included as a covariate in the model, to assist in statistically testing the effect of ELR, while blocks were included as a random variable. PROC REG was also used to validate ANCOVA model results and visually assess the fit of linear and quadratic regression models for each variable of interest. Various fit statistics, including mean square error and the predicted residual error sum of squares, were used to determine the optimal fit. Data means, mean standard error, and regression lines depicted in graphs were included to assist with data and trend interpretation, particularly with interpreting linear versus quadratic responses of variables to increasing ELR severity. Additionally, since PROC MIXED does not generate r2 values for ANCOVA mixed models, r2 values were also generated using PROC REG to aid in data interpretation. Juice free volatile composition was analyzed using JMP Pro and all metabolites were screened for significant relationships with increasing ELR severity using Pearson’s correlation coefficients. Relationships between chemical classes and ELR severity were assessed as described above using SAS. Graphs were constructed using Origin Pro v. 2022 (OriginLab Corporation).
Results
Seasonal weather conditions
Seasonal rainfall and heat accumulation (GDD) were both higher in 2018 than in 2017 (Table 1). On a monthly basis, GDD was always greater in 2018, except for in April and July; GDD during the ripening period (mid-August to harvest) was ~1.3 times higher in 2018 than in 2017. Relative to 2017, monthly cumulative rainfall was greater in 2018, except for in April, May, and July. During the ripening period, cumulative rainfall in 2018 was ~3.5 times higher than the same period in 2017, with 24% of the seasonal rainfall occurring during this period, compared to 9% in the previous year. Weather conditions around bloom and ELR treatment application in both 2017 and 2018 were warm with little rainfall (2.4 and 6.1 mm, respectively), with maximum daily temperatures reaching 33 and 32°C in 2017 and 2018, respectively, in the week following ELR applications.
Monthly ripening period (mid-August to harvest) and seasonal (April to harvest) growing degree days (GDD) and rainfall (mm) at the experimental site for 2017 and 2018.
Effects of increasing ELR on leaf area, fruit set, and cluster morphology
As expected, the amount of leaf area removed via ELR was affected by ELR in both years (Table 2). The percentage of total leaf area per shoot removed ranged from an average of 22.8% (ELR-3) to 88.4% (ELR-12) in 2017 and from 14.2% (ELR-3) to 85.6% (ELR-12) in 2018 (Table 2). The percentage of leaf area removed strongly increased from ELR-3 to ELR-8; thereafter, proportional increases of leaf area removed were much lower. Increasing ELR severity was more strongly related to main leaf area removed in both years compared to the lateral leaf area removed, and relationships between ELR severity and main and lateral leaf area removed in 2017 were stronger than those measured in 2018 (Table 2).
Main, lateral, and total leaf area (cm2), and proportion of leaf area (%) removed from Grüner Veltliner vines (n = 8), following application of a range of early leaf removal (ELR) severities (three to 12 leaves per shoot) at prebloom (E-L 19) in 2017 and 2018.
Leaf area measured at harvest was affected by ELR in both years but trends between leaf area and ELR severity were weak overall (Table 3) and much weaker than those between the leaf area removed at E-L 19 and ELR severity (Table 2). In 2017, total leaf area per shoot was 17.3 and 6.4% greater for ELR-3 and ELR-6 vines, respectively, compared to ELR-0, which was mainly the result of greater lateral leaf area of these two treatments than all the other vines. However, in 2018, only ELR-3 vines had higher total leaf area than ELR-0 (27.1%), which was explained by the greater main and lateral leaf area per shoot. Total leaf area of vines exposed to higher ELR severities (eight to 12) was between 33.9 (ELR-8, 2017) to 40.2% (ELR-12, 2017) lower than ELR-0 vines.
Main, lateral, and total shoot leaf area (cm2) at harvest in 2017 and 2018 for Grüner Veltliner vines exposed to a range of early leaf removal (ELR) severities (0 to 12 leaves per shoot) at prebloom (E-L 19).
In general, ELR effects on fruit set were stronger on basal than distal clusters (Figure 1A and Supplemental Figure 1A). There were linear reductions in fruit set with increasing ELR severity for basal clusters in both years, although the relationship was weaker in 2018 than in 2017 (pELR*YEAR = 0.0179), which in turn altered other cluster characteristics such as berry number per cluster and berry weight (Figure 1). Percent fruit set in basal clusters was comparable between seasons for ELR-0 (48.1 and 45.6%) and ELR-3 vines (41.9 and 42.2%), while it was lower for all other treatments in 2017 than in 2018. However, ELR did not affect fruit set in distal clusters (pELR = 0.177). For all vines, the number of florets measured prior to ELR was lower overall in 2018 than in 2017 (Figure 1B and Supplemental Figure 1B). As expected, the number of florets tended to be similar among all treatments in 2017. However, in 2018, vines exposed to extremely high ELR rates (ELR-10 and ELR-12) in the previous year tended to have fewer florets than all other vines. These trends were similar between basal and distal cluster, despite their divergent response to fruit-set reduction.
Relationships between basal cluster fruit set (A), floret number per cluster (B), berry number per cluster (C), average berry weight (D), and increasing early leaf removal (ELR) severity (0 to 12 leaves per shoot) applied at trace bloom (E-L 19) for Grüner Veltliner vines in 2017 (unfilled circles; n = 8) and 2018 (filled circles; n = 8). Circles indicate mean values and bars indicate the standard error, determined using PROC GLIMMIX and mixed model analysis of covariance in SAS. Regression models (2017 and 2018, solid and dashed lines, respectively) determined using PROC REG and corresponding r2 values are shown in each panel.
Similar to trends observed for fruit set, ELR affected the number of berries per basal cluster and berry weight (pELR < 0.001), but the nature of the trends between berry number and ELR severity also varied by year (pELR*YEAR = 0.0314; Figure 1C and 1D). There was a linear and negative relationship between increasing ELR severity and berry number per cluster in 2017, and reductions ranged from 16.6 (ELR-3) to 46.6% (ELR-12), relative to ELR-0 (Figure 1C). ELR severity also affected berry number per cluster in 2018, but the relationship was quadratic given that reductions in berry number, relative to the ELR-0, were measured mainly in ELR-10 and ELR-12 vines (22.4 and 32.4% reductions, respectively). In both years, the relationship between berry weight and ELR severity followed a quadratic trend, but the nature of the trends differed between years (Figure 1D). In 2017, ELR-3 to ELR-8 vines tended to have higher berry weight than ELR-0, with ELR-6 vines having the greatest berry weight (27.7% higher than ELR-0). In contrast, in 2018, all vines subjected to ELR tended to have lower average berry weight than ELR-0, with a narrow range of reduction from 1.5 (ELR-3) to 18.1% (ELR-12).
In distal clusters, ELR affected both berry number (pELR = 0.001) and berry weight (pELR < 0.001) (Supplemental Figure 1C and 1D). In both years, relationships between ELR severity and berry number were curvilinear, while there was a linear and negative relationship between increasing ELR severity and berry weight.
Effects of increasing ELR severity on vine yield parameters
In general, changes in cluster metrics induced by ELR resulted in lower yield per vine (pELR < 0.001) and lower cluster weight in ELR vines, although trends differed between years only for cluster weight (pELR*YEAR = 0.003) (Figure 2). In 2017, there was a linear, negative relationship between increasing ELR severity and yield per vine and cluster weight; however, in 2018, only high severities of ELR (ELR-8 to ELR-12) tended to alter these parameters (Figure 2A and 2B). In addition, yield per vine and cluster weights were lower overall in 2018, but the interaction was only significant for cluster weight. For instance, in 2017 and 2018, yield per vine ranged from 7.5 (ELR-0) to 3.8 kg/vine (ELR-12; 49.5% decrease) and from 5.6 (ELR-0) to 2.8 kg/vine (ELR-12; 50.2% decrease), respectively. Cluster number per vine at harvest was only affected by ELR in 2018 (pELR*YEAR = 0.008), as the relationship between cluster number per vine and increasing ELR severity was curvilinear in 2018 (Figure 2C). Only vines subjected to high ELR severities (ELR-8 to ELR-12) had reductions in cluster numbers, ranging from 13.6 (ELR-8) to 23.5% (ELR-10).
Relationships between average yield per vine (A), cluster weight (B), cluster number per vine (C), and increasing early leaf removal (ELR) severity (0 to 12 leaves per shoot) applied at trace bloom (E-L 19) for Grüner Veltliner vines in 2017 (unfilled circles; n = 8) and 2018 (filled circles; n = 8). Circles indicate mean values and bars indicate the standard error, determined using PROC GLIMMIX and mixed model analysis of covariance in SAS. Regression models (2017 and 2018, solid and dashed lines, respectively) determined using PROC REG and corresponding r2 values are shown in each panel.
Effects of increasing ELR severity on fruit ripeness parameters and sugar allocation
ELR affected juice chemistry parameters (TSS, pELR = 0.018; pH, pELR*YEAR = 0.011; TA, pELR*YEAR < 0.001), but strong relationships between these metrics and ELR severity were only apparent in 2017 (Figure 3). In the first year of the study, juice TSS and pH tended to linearly increase with increasing ELR severity (Figure 3A and 3B). However, increases were not >1 Brix for TSS (ranging from 18.9 [ELR-0] to 19.9 Brix [ELR-12]) or >0.2 for pH (ranging from 3.4 [ELR-0] to 3.6 [ELR-12]). Increasing ELR severity was linearly related to reductions in juice TA in 2017, with values ranging from 6.3 (ELR-0) to 4.7 g/L (ELR-10).
Relationships between juice total soluble solids (TSS) (A), pH (B), titratable acidity (TA) (C), and increasing early leaf removal (ELR) severity (0 to 12 leaves per shoot) applied at trace bloom (E-L 19) for Grüner Veltliner vines in 2017 (unfilled circles; n = 8) and 2018 (filled circles; n = 8). Circles indicate mean values and bars indicate the standard error determined using PROC GLIMMIX and mixed model analysis of covariance in SAS. Regression models (2017 and 2018, solid and dashed lines, respectively) determined using PROC REG and corresponding r2 values are shown in each panel.
The total sugar allocated per berry and per shoot was affected by ELR across both years (pELR < 0.001 for both parameters), but total sugar/cm2 leaf area did not vary among treatments (pELR = 0.460) (Figure 4). The relationship between sugar allocated per berry and increasing ELR severity was quadratic, but trends differed between years (Figure 4A). In 2017, total sugar per berry was higher in ELR-3 and ELR-6, both by 12.8%, relative to ELR-0, but was lower in ELR-10 and ELR-12. Comparatively, in 2018, all vines exposed to ELR severities greater than three leaves per shoot had lower sugar per berry than ELR-0. In both years, total sugar per shoot values were comparable between ELR-0, ELR-3, and ELR-6 vines, but greater ELR severities had progressively lower total sugar per shoot. Total sugar per shoot was higher in 2017 than in 2018 for all vines (pYEAR = 0.001) due to higher cluster weights and juice TSS in 2017. However, relative reductions were similar between 2017 and 2018, ranging from 5.1 (ELR-6) to 44.6% (ELR-12) in 2017 and from 2.5 (ELR-6) to 43.4% (ELR-12) in 2018.
Relationships between fruit sugar allocation metrics, including total sugar per berry (A), total sugar per shoot (B), and total sugar per cm2 leaf area (C), and increasing early leaf removal (ELR) severity (0 to 12 leaves per shoot) applied at trace bloom (E-L 19) for Grüner Veltliner vines in 2017 (unfilled circles; n = 8) and 2018 (filled circles; n = 8). Circles indicate mean values and bars indicate the standard error determined using PROC GLIMMIX and mixed model analysis of covariance in SAS. Regression models (2017 and 2018, solid and dashed lines, respectively) determined using PROC REG and corresponding r2 values are shown in each panel.
ELR did not have a consistent effect on rot incidence (pELR*YEAR = 0.001), assessed as the percentage of clusters with rot, or severity (pELR*YEAR = 0.034), assessed as the percentage of individual berries with visible rot infection within a cluster, between years (Figure 5). The only relevant trend between ELR severity and rot parameters was seen in 2017, when increasing ELR severity lowered the percentage of clusters with rot (rot incidence); however, rot incidence in 2017 was never >15% and was much lower than in 2018, when rot incidence was always >50% (Figure 5A).
Relationships between rot incidence (A) and rot severity (B) and increasing early leaf removal (ELR) severity (0 to 12 leaves per shoot) applied at trace bloom (E-L 19) for Grüner Veltliner vines in 2017 (unfilled circles; n = 8) and 2018 (filled circles; n = 8). Rot incidence represents the percentage of clusters per vine exhibiting symptoms of bunch rot at harvest, while rot severity represents the percentage of berries within clusters sampled at harvest that were affected by rot. Circles indicate mean values and bars indicate the standard error determined using PROC GLIMMIX and mixed model analysis of covariance in SAS. Regression models (2017 and 2018, solid and dashed lines, respectively) determined using PROC REG and corresponding r2 values are shown in each panel.
Effects of ELR severity on juice free volatile composition
A total of 57 unique volatile compounds were detected in Grüner V. juice at harvest using HS SPME-GC-MS. Of these compounds, 37 appeared in juice from both seasons, 14 only in 2017, and six only in 2018 (Supplemental Table 1). The identities of all compounds were validated using authentic chemical standards, except for four unidentified compounds (a tertiary alcohol, a fatty acid derivative, a phenylpropanoid, and an organic acid) and 3-ethyl-4-methylpentan-1-ol, an alcohol that remained tentatively identified using Kovats retention indices. Four compounds were negatively correlated with ELR severity in 2017: an unidentified tertiary alcohol, benzyl alcohol and benzaldehyde (both phenylpropanoids), and butanoic acid (a volatile acid); conversely, only 2,6-nonadienal, a fatty acid derivative, was positively correlated with ELR severity (pELR < 0.05; Table 4). In 2018, 3-ethyl-4-methylpentan-1-ol (an alcohol), 4-hydroxy-4-methyl-2-pentanone (a fatty acid derivative), and benzaldehyde were negatively correlated with ELR severity (pELR < 0.05; Table 4). When correlations were assessed using the two years of data combined, only benzaldehyde was negatively correlated with ELR severity (Table 4).
Relationships between increasing early leaf removal (ELR) severities (0 to 12 leaves per shoot) and concentrations of unidentified, tentatively identified, and authenticated volatile compounds detected in Grüner Veltliner juices in 2017 and 2018 using headspace solid-phase microextraction-gas chromatography-mass spectrometry. Values presented are Pearson’s correlation coefficients (r) calculated for 2017, 2018, and a combined dataset of 2017 and 2018.
When compounds were assigned to specific classes of volatiles based on their chemical structure, ELR only had a significant effect on volatile phenylpropanoids (pELR = 0.003; Supplemental Figure 2). Concentrations of volatile phenylpropanoids were lower with increasing ELR severity in both years, although linear trends were weak (r2 = 0.16 and 0.20, respectively, for 2017 and 2018). On average, fatty acid derivatives represented the largest chemical class in juice from all vines in both years (>94% of total), based on total compound concentrations.
Carryover effects of ELR severity on fruitfulness and mortality
ELR treatments imposed in 2017 did not affect the percentage of live buds (pELR = 0.680) or the number of clusters per shoot (i.e., fruitfulness) when assessed in the spring (23 May 2018) of the following year (pELR = 0.204) (data not shown). Application of ELR treatments on the same vines for two consecutive years had no effect on the mortality of primary, secondary, or tertiary buds (pELR = 0.203, 0.783, and 0.950, respectively) when measured following four consecutive days in late January and early February 2019 with minimum temperatures of −18.3, −19.3, −17.1, and −17.0°C (data not shown).
Discussion
ELR can be an effective tool for reducing cluster rot, enhancing fruit technological maturity, and influencing secondary metabolite composition (VanderWeide et al. 2021). These benefits are achieved by removing enough leaves (e.g., five to six leaves) around bloom to increase fruit zone sun exposure and induce a carbon limitation that alters cluster morphology without undesirable carryover effects on yield. However, cluster responses and carryover effects can vary across studies and cultivars. In our study, we used a wide range of ELR severities (three to 12 leaves per shoot), including extreme levels not recommended for commercial grape production. Our goal was to identify an ELR threshold for high-yielding, vegetatively vigorous Grüner V. that can maximize improvements to fruit quality and composition without undesirable carryover effects.
Increasing ELR severity progressively lowers yield, with extreme severities inducing further carryover penalties
The responses of yield parameters to increasing ELR severity followed expected trends in 2017 but not in 2018, likely in part because of indirect carryover effects (e.g., number of clusters per vine, number of florets per cluster) induced by severe levels of ELR. In 2017, the linear and negative relationship between increasing ELR severity and yield was mainly associated with a linear decrease in cluster weights. As reported in previous studies, the decrease in cluster weight in ELR-treated vines was likely caused by lower numbers of berries per cluster (Acimovic et al. 2016, Hickey and Wolf 2018) rather than by reductions in berry weight, except for ELR-12. Contrary to expectations, however, in 2018 yield was only lower in vines exposed to ELR severities of eight to 12 leaves per shoot. These reductions were related to lower cluster weights (a consequence of reduced berry numbers and berry weights) and reductions in the number of clusters per vine at harvest.
While we removed a substantial proportion of functional leaf area in both years, comparable with other ELR studies (Acimovic et al. 2016, VanderWeide et al. 2020), we found variable reductions in fruit set that may help explain the different trends observed in the two years. First, ELR induced a stronger reduction in basal cluster fruit set in 2017 (12.9 to 41.9%) than in 2018 (7.4 to 20.4%), potentially explaining why overall reductions in berry number and cluster weight in ELR vines were not as strong in 2018. Across both years, differences in basal cluster fruit set were predominantly observed in ELR-8, ELR-10, and ELR-12. Similarly, in cool-climate Pinot noir, vines exposed to a range of ELR severities (four to 10 leaves per shoot) had progressively lower fruit set with increasing ELR severity. These trends were stronger in the first of two years of the study (10.5 to 61.1% versus 20.4 to 31.6% for basal clusters in 2017 and 2018, respectively), when differences were mainly reported for vines subjected to removal of eight and 10 leaves per shoot (Acimovic et al. 2016).
Second, ELR-mediated decreases in Grüner V. yield in 2018 were in part a consequence of lower numbers of florets per inflorescence (ELR-10 and ELR-12 for both basal and distal clusters) and lower numbers of clusters per vine at harvest (ELR-8, ELR-10, and ELR-12), widely reported elsewhere (Sabbatini and Howell 2010, Risco et al. 2014, Acimovic et al. 2016, Hickey and Wolf 2018, Nicolosi et al. 2021). A reduction in cluster number may be attributed to adverse effects of ELR-mediated carbohydrate limitation on inflorescence initiation (Noyce et al. 2016b). We did not detect a clear effect of ELR severity on fruitfulness when assessed in the spring of 2018, prior to shoot-thinning, while we noted a reduction in cluster number per vine later at harvest. However, we might not have fully captured the effects of ELR on fruitfulness, as we only assessed fruitfulness on two random cordons per experimental unit, while the cluster number at harvest was calculated for each experimental vine. Nevertheless, while we hypothesized that a vegetatively vigorous cultivar like Grüner V. would require a higher ELR severity to achieve targeted fruit set and yield reductions, these negative carryover effects suggest that more severe ELR severities may not actually be necessary for cultivars like Grüner V.
Additionally, little work has evaluated how ELR affects the dynamics of soluble carbohydrates important for bud freeze tolerance during dormancy, another mechanism that can affect the vine crop potential. Previous work on ELR in Grüner V. indicated no negative effect of ELR on bud freeze tolerance when five leaves per shoot were removed (Smith and Centinari 2019). We did not assess bud freeze tolerance but rather bud mortality, following two successive years of ELR application and a severe winter freeze event that caused significant bud mortality, regardless of ELR treatments. While we did not assess fruitfulness in the year following 2018 as well, these results nevertheless suggest that reductions in yield are tied to the influence of ELR-mediated carbon limitation on floret and cluster number, at least in this case.
Interaction between floret number and carbon limitation affected berry responses to increasing ELR severity
In 2018, the lack of a clear fruit-set response to increasing ELR severity may have been caused by the overall lower number of florets per inflorescence in basal and distal clusters measured prior to ELR application. We reason that carbon competition imposed by removing three to eight leaves per shoot may not have been strong enough to limit fruit set in Grüner V. inflorescences with relatively low numbers of florets, explaining why ELR effects on yield parameters such as berry number and cluster weights were not more pronounced in these vines (ELR-3 to ELR-8) in 2018. However, removing 10 to 12 leaves per shoot likely imposed a sufficient degree of carbon limitation to reduce berry number per cluster in both basal and distal clusters at harvest in 2018. Additionally, these same vines (ELR-10 and ELR-12) appeared to have lower floret numbers in both clusters at E-L 19 in spring 2018 than all the other treatments, suggesting that the application of extreme ELR severities the preceding season may have had an adverse effect on inflorescence primordia development prior to a second, successive application of ELR. The initialization of basal floral primordia is reported to occur in latent buds of Chardonnay at stage E-L 18 (Noyce et al. 2016a), about the same time as when we applied ELR (E-L 19) in this study. Both the size and number of inflorescence primordia can be negatively affected by carbon limitation induced by defoliation (Noyce et al. 2016b). We did not measure carbohydrate concentrations in perennial tissues to directly assess if severe ELR application (e.g., ELR-10 and ELR-12) led to a sustained carbon limitation scenario that lasted throughout the season and could have negatively affected primordial development. However, this trend in floret number was also reported for Pinot noir subjected to ELR severities of eight and 10 leaves per shoot (Acimovic et al. 2016).
The responses of berry weight to ELR also diverged between years in basal clusters, highlighting two different changes in carbon dynamics potentially induced by ELR. In the first year, increasing ELR severities from three to 10 leaves were associated with linear decreases in berry number per cluster, and increased or comparable berry weights relative to ELR-0, as the remaining berries likely received a greater allocation of resources (Poni and Bernizzoni 2010, Hed et al. 2015). Comparatively, this compensation effect was not present in the second year when the clusters overall had fewer but heavier berries. This was likely due in part to all vines having a lower number of florets per inflorescence prior to a second year of ELR application, regardless of any carryover effects from ELR application the previous year. Consequently, higher ELR severity and the potential carbon competition between berries appeared to steadily lower berry weights instead (Gatti et al. 2012). Considering both years, it appears that for Grüner V., the effects of ELR on berry weights of basal clusters were in part a function of the original floret number, since increasing carbon limitation due to the same ELR severities appeared to have different effects when the number of competing florets differed, while for distal clusters, increasing ELR severity led to consistent reductions in berry weight.
Increasing ELR severity reduced rot and improved fruit ripeness in year one only
In humid grapegrowing regions, a major goal of ELR is to reduce cluster rot (VanderWeide et al. 2021); in our study, however, the efficacy of ELR for rot reduction in Grüner V. was limited to a single season (i.e., 2017). In 2018, high rainfall during the ripening period (212 mm, 24% of seasonal rainfall) promoted elevated levels of rot incidence (>60% of clusters) and severity (>20% of berries), likely overshadowing any positive influence of ELR on fruit rot. We did not separately quantify B. cinerea and non-Botrytis rots, but high rainfall and warm temperatures during the ripening period likely facilitated high levels of sour rot. Previous assessment of ELR in Pinot grigio across two years with variable precipitation also indicated that ELR was more effective at mitigating fruit rot development in a drier year, with positive effects on both Botrytis bunch rot and sour rot (VanderWeide et al. 2020). To date, there are contrasting results regarding the relationship between ELR and fruit rot development. Reported reductions in fruit rot and disease pressure due to ELR have been attributed to a variety of factors, including reductions in berry number that can reduce cluster compactness (Poni et al. 2006, Intrieri et al. 2008), improved sunlight exposure and air movement (Vogel et al. 2020, Hed and Centinari 2021), and lower retention of floral debris (Hed and Centinari 2018). However, ELR can be equally or less effective at reducing fruit rot than leaf removal applied at later phenological stages, including at or after fruit set (Smith and Centinari 2019, Vogel et al. 2020). In some cases, this may be related to ELR having no effect on cluster compactness (Hed and Centinari 2018). We did not measure cluster compactness, but fruit from ELR-treated vines visually appeared to have looser clusters than ELR-0 (personal observation). Additionally, it was out of the scope of this study to compare the timing of leaf removal. The choice of ELR versus leaf removal performed at later growth stages in Grüner V. would ultimately depend on grower goals and whether crop management (i.e., lower yield) is desired alongside rot control, for example.
Reported improvements in fruit maturity due to ELR have mainly encompassed an increase in TSS at harvest (reviewed in VanderWeide et al. 2021), but in our study, ELR only improved juice TSS in the first year and increases were relatively minor. In 2017, fruit TSS increased incrementally with ELR severity up to 1 Brix (ELR-12). Comparatively, Semillon vines exposed to ELR of five leaves per shoot had higher TSS at harvest in both years of a two-year study (2.8 and 1.4 Brix), relative to control vines (Alessandrini et al. 2018). However, other studies also reported contrasting effects of ELR on juice TSS between years, as in our study. For example, juice TSS at harvest of berries from Pinot grigio vines exposed to mechanical ELR of six leaves per shoot was higher only in one of two years (by ~2 Brix), relative to control vines (VanderWeide et al. 2021). Similarly, increasing severities of ELR in Pinot noir led to an increase in TSS (0.8 to 3.2 Brix) only in one year of two years of study as well (Acimovic et al. 2016). Higher pH in ELR-treated vines has been reported in several studies (Poni et al. 2008, Gatti et al. 2012, Acimovic et al. 2016), similar to our work. In our study, however, juice pH for more severe ELR (ELR-8 through ELR-12) in 2017 was close to the upper threshold of pH recommended for wine microbial stability (i.e., pH 3.7) and would require enological adjustment (Bartowsky 2009, Waterhouse et al. 2016). It should be noted that freezing the fruit prior to analysis may also have contributed to the overall high juice pH that we measured. In sum, given the minimal increase in TSS and the risks for stability due to high juice pH, growers of Grüner V. may be willing to accept an ELR severity that yields a modest improvement in juice TSS while minimizing undesirably high pH (i.e., ELR-6 or ELR-8).
An increase in TSS due to ELR in 2017 may have been a function of increased sugar allocation to the berry in ELR-treated vines. Greater improvements in TSS due to ELR reported elsewhere have been attributed to different factors that can increase the allocation of sugar to fruit, including greater carbon assimilation rates (Poni et al. 2008) and/or increased lateral leaf area growth during the season, resulting in higher leaf area-to-yield ratios (Palliotti et al. 2011). In our study, however, only ELR-3 in both seasons and ELR-6 in 2017 had greater lateral leaf area than ELR-0. Grüner V. vines used in this study were trained to a double-Guyot system with vertical shoot positioning, necessitating implementation of two hedging passes to reduce vine vegetative growth. However, while hedging may have limited extensive leaf area regrowth after ELR, the lateral leaf area on a per shoot basis measured at harvest was greater for all ELR-treated vines in 2017 than what was reported for potted and field-grown Sangiovese (Poni et al. 2006, Intrieri et al. 2008), field-grown Graciano (Tardaguila et al. 2010), and field-grown Lambrusco (Poni et al. 2009) vines, all of which had increased TSS in ELR-treated vines and had been hedged once. Therefore, different weather conditions (sunlight and temperature) observed between our study site and previous work conducted in Mediterranean climates might better explain the smaller effects of ELR on sugar accumulation in fruit reported here.
While we did not measure carbon assimilation, we reason that improvements in juice TSS may be a result of fewer berries competing for sugar, as reductions in fruit mass appeared to parallel those in leaf area with increasing ELR (Poni et al. 2006, Intrieri et al. 2008). This may partly explain why no improvement in juice TSS was measured in 2018, a year when ELR did not affect berry number per cluster except in ELR-10 and ELR-12. In 2018, high rainfall during the ripening period (i.e., 212 mm) may also have potentially masked any effect from ELR, if the persistently high levels of humidity and water on berries led to direct water uptake and a dilution effect in the berry (Keller 2015). However, in both years the total sugar per shoot was lower in ELR-10 and ELR-12, relative to ELR-0, suggesting that lower carbon allocation may be a direct influence. Comparatively, the absence of any increase in TSS in ELR-10 and ELR-12 in 2018 may also be masking an upward trend in TSS observed for lower ELR severities (i.e., ELR-6 and ELR-8) relative to the non-defoliated vines, since the chosen statistical analysis assessed the overall trend across all treatment vines.
Increasing ELR severity mainly impacted juice volatile phenylpropanoids
Despite the significant changes to yield parameters and fruit-zone microclimate (data not shown) caused by increasing ELR severity in our study, ELR did not cause strong, positive responses in volatile concentrations, as we had hypothesized. Unexpectedly, ELR instead only affected the total concentration of phenylpropanoids in Grüner V. juices in both years, mainly due to ELR having a consistent, negative effect on benzaldehyde, a volatile phenylpropanoid that can positively contribute a “marzipan” or “cherry” aroma to juices and wines (Waterhouse et al. 2016). Effects of leaf removal on benzaldehyde have been investigated in other cultivars but with contrasting findings. For example, benzaldehyde concentrations in Semillon juices or wines were not affected by ELR of five leaves per shoot (Alessandrini et al. 2018), but in Cabernet Sauvignon berries, benzaldehyde concentrations were increased by leaf removal performed at veraison (He et al. 2020). Furthermore, as concentrations of another phenylpropanoid, benzyl alcohol, were lower in Semillon wines from ELR-treated vines in one of two years, ELR may impact phenylpropanoids more broadly (Alessandrini et al. 2018, and also found in the present study in 2017). Volatile phenylpropanoid synthesis is linked to other biosynthetic pathways responsible for red grape pigmentation, which can be affected by ELR-mediated changes in fruit microclimate and ripening in red grapes (VanderWeide et al. 2021). In Sauvignon blanc, ELR increased the concentrations of skin tannins and phenolics in fruit for only a single year of a two-year study (Komm and Moyer 2015). While we did not assess phenolic composition, we noticed a perceptible visual difference in the color of fruit from ELR-treated vines in 2017 (personal observation), likely a consequence of increased cluster solar exposure that may signify increased phenolic concentrations (Ristic et al. 2007). Despite this, the reports of negative effects on benzaldehyde and benzyl alcohol suggest that ELR may affect volatile and nonvolatile phenolic compounds differently. Additionally, it is unclear to what degree an ELR-mediated shift in phenolics could have on Grüner V. wine quality and consumer sensory perception.
While ELR appeared to broadly affect phenylpropanoids, it had minimal effect on other volatiles that are sensitive to microclimatic changes, including terpenes and fatty acid derivatives (Alem et al. 2019, He et al. 2020). Hotrienol, which has been previously detected in Grüner V. juice but not in the resulting wine (Keller 2020), was the only terpene negatively affected by ELR, in 2017 only. Terpenes may play a limited role in influencing the composition of Grüner V. wines, unlike aromatic white grapes (such as Riesling) that comparatively have much higher terpene concentrations. Although the terpenes of aromatic cultivars are sensitive to microclimatic alterations (Reynolds et al. 1996), other studies reported variable effects of ELR on terpene content. Total terpene content was higher in Riesling juices due to ELR of four to five leaves per shoot in only one out of two years of the study, while no effect was seen in Sauvignon blanc juices over two consecutive years (Komm and Moyer 2015). Similarly, in Semillon, ELR of five leaves per shoot did not affect total free terpene content; only select terpenes (e.g., diendol and trans-pyran linalool oxide) were affected in a single year out of two years (Alessandrini et al. 2018).
As changes in the concentrations of some compounds in response to greater fruit-zone exposure have been reported (He et al. 2020), we hypothesized that ELR could affect fatty acid derivatives. However, we found that ELR had a minimal and inconsistent effect on Grüner V. juice compounds. Consistent with previous findings that C6 alcohols and aldehydes (believed to be responsible for “green” aromas [Robinson et al. 2014]) are present in high concentrations in grape juice at harvest (Kalua and Boss 2010), in our study, fatty acid derivatives made up most of the compounds present in Grüner V. juice. Despite their prevalence, ELR affected only two of 19 identified compounds ((E,Z)-2,6-nonadienal and 4-hydroxy-4-methyl-2-pentanone), and only in a single year for both compounds. Similarly, the concentrations of various free and glycoside volatile alcohols in musts from ELR-treated Semillon vines were higher than those in control vines (Alessandrini et al. 2018), while there were mixed effects on volatile acids and aldehydes in Riesling and Sauvignon blanc juices (Komm and Moyer 2015). While we imposed a wider range of ELR that resulted in more severe microclimatic alteration and, most likely, greater sunlight exposure, in our study, ELR appeared to have less of an impact on these compounds. In contrast to other studies, we blended the berries of Grüner V. used for volatile composition analysis and performed the analysis on the unfiltered juice. In contrast to other studies (Alessandrini et al. 2018), we did not treat the juice with enzymes to release glycosidically-bound volatiles prior to analysis. However, these compounds are released upon wounding of plant tissue (Schwab et al. 2008), and it is also likely that the blending process released some proportion of bound volatiles. If this is the case, this might explain the high concentrations of fatty acid derivatives measured here and, contrary to our hypothesis, suggests that ELR has a minor effect on chemical components other than phenylpropanoids in Grüner V. juice.
Conclusion
In the first year of this study, grapevines exposed to increasing ELR severities had gradually lower yield with less fruit rot and greater fruit maturity, achieving the two major goals of ELR: yield reduction and fruit quality improvement. Comparatively, vine responses to ELR in the second year of this study diverged from trends observed in the previous year; yield reductions were only documented in high ELR severities (eight to 12 leaves per shoot), and regardless of ELR severity, unfavorable weather conditions during ripening facilitated high levels of rot severity and incidence. Of the volatile compound classes investigated, only volatile phenylpropanoid concentrations were negatively affected by increasing ELR severity, with a consistent negative effect on benzaldehyde, suggesting a limited effect of ELR on juice free volatile composition in Grüner V. overall. Based on our results, if crop reduction is desired, application of medium ELR severities (six to eight leaves per shoot) may be optimal for high-yielding, vegetative Grüner V., as the yield values of ELR-6 and ELR-8 vines across years (~4.0 to 6.5 kg/vine) will be economically viable to Grüner V. growers. Improvements in fruit quality (namely, lower rot) were also documented in seasons without atypically high precipitation. There were no carryover effects on inflorescence floret number for these severities following one year of ELR application, but in the second year of the study, ELR-8 vines had fewer clusters per vine (17% less, or six clusters per vine) than ELR-0. However, if these ELR severities are repeatedly applied on the same vines over consecutive seasons, we cannot exclude the possibility that negative effects on floret and cluster number may occur.
Supplemental Data
The following supplemental materials are available for this article in the Supplemental tab above:
Supplemental Table 1 Validated volatile compounds identified in Grüner Veltliner juices in 2017 and 2018 using headspace solid-phase microextraction-gas chromatography-mass spectrometry. Compounds were validated by referencing calculated linear retention indices (LRIcalc) with those from published literature (LRIlit) and via authentic standards (LRIstd).
Supplemental Figure 1 Relationships between distal cluster fruit set (A), floret number per cluster (B), berry number per cluster (C), average berry weight (D), and increasing early leaf removal (ELR) severity (0 to 12 leaves shoot) for Grüner Veltliner vines in 2017 (unfilled circles; n = 8) and 2018 (filled circles; n = 8). Circles indicate mean values and bars indicate the standard error determined using PROC GLIMMIX and mixed model analysis of covariance in SAS. Regression models (2017 and 2018, solid and dashed lines, respectively) determined using PROC REG and corresponding r2 values are shown in each panel.
Supplemental Figure 2 Boxplots showing the total concentration of volatile compounds summed by chemical group in juices of Grüner Veltliner in 2017 (dark grey box; n = 6) and 2018 (light grey box; n = 6) in relation to increasing early leaf removal (ELR) severity (0, 6, 8, 10, and 12 leaves per shoot). Within the interior of each box, interior horizontal black lines represent the median, and the small interior box represents the mean, while the upper and lower limits of the vertical box exterior represent the 25 to 75% range of the data. The whiskers of each box represent a range corresponding to 1.5 times the interquartile range. Black diamonds above or below the box indicate single datapoints that are designated as outliers, using the interquartile method.
Footnotes
This work was supported by the Pennsylvania Wine Marketing Research Board Program and by the USDA NIFA Federal Appropriation under Projects PEN0 4794 and PEN0 4792 (Accession numbers 7003432 and 7002577, respectively). We express gratitude toward Don Smith, Colton Craig, and Deanna Homan for technical and field assistance and Andrew Poveromo and Marielle Todd for lab assistance. The authors thank Dr. Charles Zaleski, MD, at Fero Vineyards and Winery for providing and maintaining the vineyard experimental site.
Harner AD, Smith MS, Keller ST, Hopfer H and Centinari M. 2024. Identifying an early leaf removal threshold for Grüner Veltliner, a high-yielding, high-vigor cultivar. Am J Enol Vitic 75:0750005. DOI: 10.5344/ajev.2024.23021
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 April 2023.
- Accepted January 2024.
- Published online March 2024
This is an open access article distributed under the CC BY 4.0 license.