These treatments are provided visually in Figure 1.In the second experiment, individual grapes from different cluster positions were collected from two cultivars grown in a commercial vineyard in Oakville, CA in 2017. Cabernet Sauvignon grapevines and Petit Verdot grapevines were 21 and 9-years old, respectively. The exposure of each individual grape was estimated with fish-eye lens photography from the grape perspective pointing the zenith. The images were processed in R . After applying a thresholding condition to the blue channel of all images, they were converted into binary pixels . Thus, the percent of binary pixels capturing the sky was used to calculate the percentage of canopy porosity as reported previously . Then, those berries were collected at harvest, and their flavonoid content was analyzed with reversed-phase high performance liquid chromatography.The experiment was conducted in 2019 in Oakville, CA with row orientation NW-SE. The vineyard was spaced 2 m × 2.4 m with Cabernet Sauvignon grapevines on 110R root stock. The grapevines were trained to a vertically shoot-positioned system with a cordon height 96 cm above vineyard floor, trained to a bilateral cordon, and pruned to 1-bud spurs. Plants were irrigated weekly with 2-drip emitters per vine, with the capacity to deliver 3.8 L of water per hour. The experiment was designed as a randomized complete block with three canopy management practices: removal of 5 to 6 basal leaves on the NE side ; thinned to 24 shoots per vine ; and a combination of LR and ST and an untreated control ,square plant pots with four replicates each consisting in 5 grapevines, 3 of which were sampled and the 2 on distal ends were treated as border plants.
The ST and LR treatments were applied on 11 June 2019. Harvest commenced when the berry TSS reached to ca. 24°Brix on 23 September . The sampling time points were as follows: 2 weeks before veraison , veraison , 2 weeks after veraison , 3 weeks after veraison , 5 weeks after veraison , and harvest , were chosen to cover the response of the berry metabolism to cultural practices and the concomitant increase in exposure.Leaf area index was measured on 21 June to characterize grapevine canopy growth and converted into leaf area on by a smartphone based program, VitiCanopy, coupled with an iOS system . The gap fraction threshold was set to 0.75, extinction coefficient was set to 0.7, and sub-divisions were 25. A “selfie-stick” was used for an easy access to place the device about 75 cm underneath the canopy. The device was positioned with the maximum length of the screen being perpendicular to the cordon, and the cordon being in the middle of the screen according to previous work . In each experimental unit, three images were taken to capture half canopy of each vine, and analyzed by the software. The relationship between leaf dry mass and area was determined on a sub-sample of leaves of different sizes using a leaf area meter . Total leaf area was calculated by defoliating one grapevine per treatment replicate after harvest and using the regressive relationship between leaf dry mass and leaf area. At harvest, clusters were manually removed, counted, and weighed on a top-loading balance. Leaf area to fruit ratio was calculated by dividing leaf area with crop weight. Dormant pruning weight was collected during the dormant season ; and crop load was calculated as the ratio between yield per vine and the pruning mass of each vine. Labor operations costs and gross income per hectare were calculated based on yield and net returns per hectare and methods presented elsewhere . Anthocyanin productivity was calculated as reported by Cook et al. .
At each sampling point and experiment, 55 berries were randomly collected from the middle of each treatment-replicate and kept on ice until they were measured. Berries were weighed, and mean berry mass was determined as the average mass of the counted berries. These berries were used to determine the total soluble solids , the pH, and the titratable acidity . TSS was measured as °Brix, with a digital refractometer . The juice pH and TA was determined with an autotitrator using sodium hydroxide to titrate to an end point of pH 8.3, and it was expressed as g•L−1 of tartaric acid.For each sampling point in each experiment, 20 berries were collected, gently peeled, and berry skins were freeze-dried . Dried tissues were ground with a tissue lyser . Fifty mg of the resultant powder was extracted in methanol: water: 7 M hydrochloric acid to simultaneously determine flavonol and anthocyanin concentration and profile as previously described Martınez-Lüscher et al. . Briefly, extracts were filtered and analyzed using an Agilent 1260 series reversedphase high performance liquid chromatography system coupled to a diode array detector. Separation was performed on a reversed-phase C18 column LiChrospher® 100, 250 mm × 4 mm with a 5-µm particle size and a 4-mm guard column of the same material at 25°C with elution at 0.5 ml per minute. The mobile phase was designed to avoid co-elution of anthocyanins and flavonols consisted in a constant 5% of acetic acid and the following gradient of acetonitrile in water: 0 min 8%, at 25 min 12.2%, at 35 min 16.9, at 70 min 35.7%, 65% between 70 and 75 min, and 8% between 80 and 90 min. The identification of flavonoid compounds was conducted by determining the peak area of the absorbance at 280, 365, and 520 nm for flavan-3-ols, flavonols and anthocyanins, respectively. Identification of individual flavan-3-ols, anthocyanins, and flavonols were made by comparison of the commercial standard retention times found in the literature. Commercial standards of epicatechin, malvidin-3-O-glucoside, and quercetin-3-Oglucoside were used for the quantification of flavan-3-ols, anthocyanins, and flavonols, respectively.
The determination of proanthocyanidins was performed using an Agilent HPLC-DAD after an acid catalysis in the presence of excess phloroglucinol , with minor modifications described in Martınez-Lüscher et al. .The 3-isobutyl-2-methoxypyrazine was quantified by a stable isotope dilution assay using headspace solid phase microextraction coupled to a gas chromatograph and a mass spectrometer as described Chapman et al. and Koch et al. with some modifications. Briefly, 20 berries per treatment-replicate from Experiment 3 were randomly collected from the clusters of three vines in the middle of each treatment-replicate on both side of the canopy, by cutting the pedicel with a pair of scissors and frozen at −80°C until analysis. Pedicels were removed by hand and berries were placed in 50 ml conical tubes. 10 ml of pure water and 100 ml of deuterated IBMP isotope were added into the tube. Then, samples were ground with a tissue homogenizer Power Gen 1800D and centrifuged at 3000 rpm for 10 min. 10 ml of the supernatant was pipetted into 20 ml SPME vials containing 3 g of sodium chloride.Statistical analyses were carried out using the R-Studio version 3.6.1 for Windows. All data were subjected to Shapiro-Wilk’snormality test . Correlations between variables were calculated with the Pearson’s test by using the same software. Segmented regression analysis was used to determine the point of inflection the in the relationship between increasing exposure and the berry skin anthocyanin and flavonol content with “segmented” 0.5-0.3 R package . Data were normally distributed and, subsequently, were submitted to an analysis of variance to assess the statistical differences between the treatments applied in each experiment performed. Means ± standard errors were calculated, and when the F value was significant , a Duncan’s new multiple range post hoc test was executed using “agricolae” 1.2-8 R package . When data were not normally distributed, a Kruskal-Wallis test was conducted. Percentage data were transformed according to the suggestion of the most likelihood test,plastic pots for planting into arcsine root square before ANOVA or Kruskal-Wallis tests.The growing season of 2017 was warmer and drier compared to the reference data for the same period within the last 20 years . Thereby, average daily temperature was 4°C higher and rainfall was 18 mm less. Grape berry mass differed significantly depending on the degree of exposure . Overexposed berries were the smallest due to overexposure resulting in dehydration thereby reducing berry mass. Neither total soluble solids nor titratable acidity changed regardless of the degree of exposure to which berries were subjected. However, the juice pH of the Exp+ Deg+ and Exp+ Deg++ berry must was greater compared to Exp− and Exp+ Deg− berries. Berry skin flavonoid content and composition were also affected by the degree of exposure . The berry anthocyanin content of Exp− was similar to Exp+ Deg−. However, overexposed berries resulted in berry anthocyanin content that was 70% and 90% lower when compared to the Exp− berries. Grape berry exposure to solar radiation not only affected the anthocyanin content but also modified the ratio between the tri- and di-substituted anthocyanins leading to a less stable profile in all treatments with exposed berries. Likewise, berry skin flavonol content and composition were strongly affected by the degree of exposure to solar radiation. Therefore, in Exp+ Deg− flavonol content was two-fold greater than Exp−, albeit they abruptly decreased in overexposed grapes where flavonol content was 25% and 50% lower when compared to Exp− berries. Furthermore, in overexposed berries the proportion of kaempferol and quercetin significantly increased while the proportion of myricetin decreased. Regarding proanthocyanidins in berries, mild exposure did not affect their content in Exp+ Deg− compared to Exp− berries.
However, greater solar exposure decreased proanthocyanidin content in berries but to a lesser extent compared to Exp−. Finally, the content of flavan-3-ols was severely reduced in Exp+ Deg++ berries .The analyses performed on single berries from two varieties confirmed the obtained response in anthocyanins and flavonols in Cabernet Sauvignon . Thus, exposure affected the accumulation/degradation of these flavonoids. Exposed berries from the East side of the canopy decreased 8%and 36% of the anthocyanin content in Cabernet Sauvignon and Petit Verdot, respectively. Thus, Petit Verdot seemed to be more sensitive to higher level of solar exposure and degraded anthocyanins. Overexposed berries of Cabernet Sauvignon resulted in an 87% decrease of the berry skin anthocyanins when compared to the interior berries . Berry skin anthocyanins and increasing exposure showed a significant trend below the 22% of kaempferol . Conversely, analysis of the segmented regression on Petit Verdot berries did not show a clear trend below the 3.2% of Kaempferol and after the point of inflection, anthocyanins started to degrade . Regarding flavonol content, no differences were observed between cultivars . Conversely, when exposure increased to ca. 60% the content of flavonols in exposed berries of both canopy sides and in both cultivars; the overexposed berries had the lowest flavonol content . Thus, our data revealed a strong positive relationship between the berry skin flavonols and the percentage of kaempferol until 8.6% of kaempferol proportion for Cabernet Sauvignon and 7.2% Petit Verdot . However, beyond these thresholds, flavonols started to degrade, and there was an indirect relationship between the flavonol content and the percentage of kaempferol for both cultivars, this relationship being significant only for Cabernet Sauvignon .The weather conditions during the execution of this experiment were highlighted by greater maximum daily temperatures when compared to the reference period . This was more prominent during the driest months . Moreover, global solar radiation received at the experimental site was to ca. 200 W m−2 greater than the total solar radiation recorded within the reference period . The combinatory effect of LR and LT treatments caused a 58% reduction of LAI and a 45% increase of canopy porosity . However, neither leaf area nor pruning mass showed significant differences between treatments. On the other hand, yield components were mostly affected by the shoot thinning treatments . Thus, shoot thinned vines showed lower number of clusters, yield, and Ravaz Index , and increased leaf area to fruit ratio per vine as expected. The extent of yield reductions was 55% and 47% for ST and LRST vines, respectively . Berry mass was not significantly affected by canopy management practices during the berry ripening although vines subjected to LRST tended to result in smaller berries . The most influential effects observed on berry chemistry were due to shoot thinning treatments . Therefore, shoot thinned vines had greater total soluble solids and lower titratable acidity from mid-ripening to harvest. However, no significant effect was observed on the must pH . Shoot thinned grapevines had higher anthocyanin content at veraison . However, we did not measure any changes to anthocyanin content at harvest as affected by the canopy management practices applied.