Introduction

Global surface temperature is expected to rise in the future, and a range of scenarios based on greenhouse gas emissions have been developed to predict the possible extent of these temperature changes (Intergovernmental Panel on Climate Change, 2014). In the lowest emission scenario, the temperature will increase up to 1.7 °C by the end of the century. In the highest emission scenario, reflecting present day emissions, a maximum increase of 4.8 °C is predicted (Cubasch et al., 2013). Global warming influences agriculture, including viticulture. Riesling is a grapevine considered best suited to cooler climates and is the most planted grape variety in Germany. Therefore, it is important to know the influence of abiotic factors on berry composition, and to understand how viticulture practices can be adapted to cope with rising temperatures.

Defoliation is used in viticulture to improve the canopy microclimate, with the maintenance of berry health and improving berry composition being two of the main underlying objectives (Smart and Robinson, 1991). There are several studies which have investigated the influence of light on fruit composition. Recent studies with Riesling, Friedel et al. (2015) examined the influence of defoliation and shading treatments for maturity parameters and phenols. Defoliation changed the composition of flavonoids, amino acids and malic acid, but sugar accumulation was not significantly affected by changes in leaf area to fruit weight ratio. Komm and Moyer (2015) investigated the influence of fruit-zone leaf removal at prebloom, bloom and four weeks postbloom on fruit quality and phenols in berry skins. They found that leaf removal did not influence harvest total soluble solids (TSS), titratable acidity (TA) and pH, also phenolics in berry skin were not altered when the fruit-zone was defoliated before or up to four weeks postbloom. Neither study investigated the temperature effect when decoupled from the influence of light. Furthermore, most leaf removal research has been conducted with red varieties, particularly in relation to the effects on anthocyanins. Most studies regarding the effect of temperature on grapevine berry composition, also examined red varieties.

Bonada and Sadras (2015) published a review of methods to investigate the temperature effect on grapevines. Direct (comparison in time or space) and indirect (greenhouse/growth chambers or field experiments) methods were compared and the response of the berry composition to temperature were summarized. Gouot et al. (2019) published a review about grape berry flavonoid responses to high temperatures. It was concluded that changes in flavonoid composition correlates with temperature, caused by exceeding the threshold of biosynthesis. The duration of temperature and different periods of higher temperatures have been less well studied. Most critical experimental parameters in the reviewed studies were phenological stages, followed by day/night temperature regimes. Blancquaert et al. (2019) summarized the effect of abiotic factors on phenolic compounds in grape berry and concluded that flavonol concentration increased with higher light exposure as well as with water deficit. Exclusion of solar UV radiation decreased the flavonol concentration, while temperature had no effect on flavonols. As a number of these recently published reviews highlight, the influence of temperature on berry composition is of considerable importance for wine grape production.

In one of the more targeted studies, Spayd et al. (2002) separated the effect of sunlight and temperature on maturity parameters, anthocyanins and flavonols in Merlot berry skin by utilising different sampling techniques, UV-absorbing/transmitting foil and air blowers. For Riesling, which has not had interactive temperature and light effects examined in such detail, the temperature effect with a direct method (comparison in time) was investigated for TSS and TA by Duchêne and Schneider (2005), Bock et al. (2011), Urhausen et al. (2011) and Vršič and Vodovnik (2012). Jackson and Lombard (1993) reviewed the influence of temperature, light and water status on TSS, organic acids, colour compounds and aroma precursors, and the process of photosynthesis, which are all driven by enzymes. Riesling is known for its balance between sweetness and acidity, higher temperatures reduce titratable acidity due to malic acid reduction (Lakso and Kliewer, 1975). TSS was increased by warmer temperatures (Mira de Orduña, 2010) while extreme high temperatures have been shown to decrease sugar accumulation and ripening (Greer and Weston, 2010). Phenols cause browning reactions in white wine (Singleton, 1987) and lead to haze formation with proteins (Batista et al., 2010; Ferreira et al., 2001). Phenols are also associated with a bitter taste and astringency (Arnold et al., 1980; Singleton et al., 1975).

The aim of this study was to separate the effect of light and temperature and to investigate their influence on maturity parameters and phenols. An open heating system (OHS) was established in the vineyard based on the experience of Sadras and Soar (2009). This set-up minimizes the effects on the light conditions in the bunch zone. To investigate the effect of solar radiation, defoliations were carried out in the bunch zone at different times. The impact of the open heating system and the effect of the defoliation are both dependent on the solar radiation. The warm air of the OHS tends to distribute a constant airflow around the entire bunch whilst defoliation and variable sun exposure of individual berries will create high temperature variability between fully exposed and lesser exposed berries.

Materials and methods

2. Treatment

To heat the bunch zone passively, an open heating system (OHS) was installed at 21 days after flowering (DAF). This consisted of four 20 m long tent-like chambers, built with highly transparent greenhouse polyethylene sheet (FVG Sun Saver Clear 5-ST, 180 µm, 3-layer Coex film), each one spanning 18 treatment vines along the row with two buffer vines on either side. The outlet of the OHS was positioned in the middle of the bunch zone at 1 m above ground (Figure 1C) with an opening width between 10 cm and 20 cm. Figure 1A shows one side of the OHS and a close up of the opening placed in the middle of the bunch zone (Figure 1B). To avoid creating differences in water availability between OHS and other treatments, the chambers were rolled up during natural rainfall.

Figure 1. One side of the open heating system in the vineyard (A), close up of the opening of the open heating system in the middle of the bunch zone (B), bunch (n=67) height distribution above ground level (C).

Save View full size Expand inline Collapse inline

On both sides of the bunch zone approximately 75% of the leaves were manually removed either at pea size (DEF_E) or around véraison (DEF_L) using 18 vines per field replicate and treatment at E-L 31 (21 DAF) and E-L 35 (71 DAF) respectively (Coombe, 1995). Re-growing leaves were not removed. In the control treatment, no canopy manipulation occurred.

    

Results

Compared to the 30-year average (30-AVG; 1981-2010), meteorological data for 2016 and 2017 differed dramatically between and within the seasons (Table 1). In May and June 2017, the temperature was warmer compared to the 30-AVG and 2016, which led to an early ripening. In 2016, from véraison onwards, it was hotter compared to the 30-AVG and 2017. The mean temperature for 2016 and 2017 from April to October was the same, and 0.7 °C higher compared to the 30-AVG. Growing degree days (GDD) for April until October were 1445 in 2016 and 1442 in 2017. Regarding the monthly sunshine hours (SSH), 2017 was sunnier compared to 2016 until August. The monthly SSH were low in June 2016 compared to the 30-AVG and to 2017 (100 SSH less). In August and September 2016 conditions were sunnier, especially during aroma development in September (+90 SSH) when compared to 2017. The monthly SSH from April to October were similar for the 30-AVG and 2017, whilst in 2016 SSH were 80 h less compared to 2017. Regarding the precipitation, 2016 was wet from April to June and afterwards dry until the middle of October compared to the 30-AVG. In 2017 the beginning of the growing season was drier compared to the 30-AVG and very wet in July and August. During the harvest period the precipitation was consistent with the 30-AVG.

The cumulative SSH and the cumulative precipitation were plotted from the day of budburst, referenced to the number of days from flowering (Figure 2). The first treatments (DEF_E and OHS) were established after flowering. Budburst was on April 26 in season 2016, which is the same as the 30-AVG for this variety, whilst for the 2017 season it was advanced by 14 days to April 12. Flowering occurred on June 15 in 2016, which is a day later than the 30-AVG, and on June 5 in 2017. Sunshine hours up to flowering differed for both seasons, with 93 SSH in 2016, and 215 SSH in 2017. In the next period until DEF_L was imposed, the cumulative SSH were similar; 364 SSH in 2016 and 323 SSH in 2017. Directly after DEF_L was initiated a very sunny period until harvest occurred in 2016, with 306 cumulative SSH whilst in 2017 only 212 cumulative SSH were recorded. At the beginning of berry development (flowering until DEF_E and OHS were initiated), about 120 more SSH were recorded in 2017 compared to 2016. At the end of the season during aroma development (initiation of DEF_L (véraison) until harvest), 2016 had about 90 cumulative SSH more than 2017. Precipitation was very high shortly after budburst until approximately 50 days after flowering in 2016 (Figure 2B), afterwards the season was dry. Reversed conditions occurred in 2017 where wet conditions after flowering occurred. In summary, 2016 was wet in the beginning and sunny at the end, 2017 was sunny in the beginning and wet at the end.

Figure 2. Comparison of cumulative sunshine hours (A) and cumulative precipitation (B) at Geisenheim for the 30-year average (30-AVG; 1981-2010), 2016 and 2017 seasons. Different phenological stages are indicated below. Triangles represent the onset of the treatments.

Save View full size Expand inline Collapse inline

Figure 3A shows the diurnal temperature profile during two days at different heights within and above the OHS compared to CON on days with low (<2 SSH) or high radiation (>12 SSH). On days with low radiation there was hardly any effect on temperature within the OHS treatment compared to CON. On days with high radiation the data loggers recorded a temperature difference of up to 13 °C in the OHS compared to CON. At the opening of the OHS (100 cm) the temperature was up to 6 °C higher compared to CON. Higher temperatures in the OHS and at the opening were recorded for about eight hours. At 10 cm and 20 cm above the opening of the OHS the temperature effect lasted for about four hours with a maximum temperature difference compared to CON of 4 °C. For two periods in the developments of the fruits of each year; the intensity of radiation was grouped in low (<5 h) SSH days, intermediate (≥5 and <10 h) SSH days and high (≥10 h) SSH days (Figure 3B). The first period was after OHS installation until véraison (DEF_L) and the second period was after véraison until harvest. After véraison until harvest the number days with high SSH was more than three times higher in 2016 compared to 2017. The period before véraison is comparable for both years.

Figure 3. Temperature day courses recorded on (A) left, days with low sunshine hours (SSH, <2 h) and right, days with high SSH (>12 h) at different heights within the bunch zone. Middle height of the bunch zone and OHS opening refers to 100 cm; (B) number of days with low, intermediate and high SSH left 21 days after flowering (DAF) until véraison and right véraison until harvest in 2016 and 2017.

Save View full size Expand inline Collapse inline

Details of the canopy structure and the light conditions inside the canopy are shown in Table 2 and Table 3. In 2016, DEF_E and CON were initially different for all parameters. At the second measuring, the percentage of interior clusters (PIC), percentage of interior leaves (PIL) and leaf layer number (LLN) showed a significant difference between DEFs and CON. At the first measuring in 2017, DEF_E differed significantly compared to the other treatments for PIC, PIL and LLN. Percentage of gaps (PG) in DEF_E was two to three times higher compared to the other treatments. After DEF_L was applied, a significant difference was observed between the treatments for all parameters. Regarding PIC and LLN, all treatments differed significantly between each other. DEF_E and DEF_L showed the lowest and almost the same values for PIL. In both years the defoliation treatments showed significant differences to CON. In 2017, CON and OHS did not show different conditions at the first measuring. At the second measuring there was a difference between CON and OHS for some parameters but substantially lower than the differences caused by the defoliation treatments.

No significant difference in solar radiation was found between treatments in the first sampling period (Table 3). Although DEF_E showed the highest amount of daily solar radiation compared to the other treatments. At the second time point a significant difference was found between DEF_E and CON. No significant difference was found for DEF_E and DEF_L or DEF_E and OHS. Comparing the means of the treatments, a clear difference between DEF_E and the other treatments occurred. At the third time point, shortly after DEF_L was applied, and at the fourth time point, solar radiation in DEF_L was significantly different to CON and OHS. DEF_E did not differ to the other treatments as only a single defoliation was applied.

Table 4 summarizes all harvest parameters for 2016 and 2017 at the end of the growing season. The influence of the treatments was significant for TSS concentration in 2016, but in 2017 no significant difference was recorded. At the beginning of véraison the berries of OHS had a higher concentration of TSS and a lower concentration of acids (Figure 4A). Compared to all other treatments, the ripening process was initially advanced for the OHS treatment, although the effect was greater in 2017 (Figure 4B). Prior to harvest the sugar accumulation rate of the OHS-treatment declined and was lower compared to the other treatments. In both years the lowest concentration in TSS was found for the OHS treatment at the last sampling date.

The glucose concentration and the fructose concentration were both significantly decreased by OHS compared to the other treatments in 2016; in 2017 no significant effects were observed. Nevertheless, in both years the ratio of glucose to fructose was significantly influenced by OHS (Table 4), the concentration of glucose decreased compared to the fructose concentration (Figure 4C,D). The malic acid concentration decreased faster in the OHS compared to the other treatments (Figure 4A,B), which caused a significant influence on the ratio of malic and tartaric acid (Table 4). In 2017, the warm and sunny period at the beginning of the season and at the installation of the OHS caused an early ripening and an intense heating effect. The malic acid concentration was approximately 4.5 g/L less at véraison (determined for translucent yellow green berries of CON) and at harvest approximately 2 g/L lower in OHS compared to the other treatments (Figure 4B). In 2016 the decline in the malic acid concentration was smaller (0.5 g/L) between OHS and the other treatments and the difference did not change during the whole season (Figure 4A).

In both years, DEF_L had the lowest concentration of tartaric acid (Table 4). In 2017, there was a significantly lower tartaric acid concentration for DEF_L compared to the other treatments. In both years, CON showed the highest concentration of shikimic acid. In 2016, the concentration of shikimic acid was the lowest for DEF_L, whilst OHS was slightly higher. In 2017, OHS and DEF_L also showed low concentrations, while OHS had the lowest concentration of shikimic acid compared to the other treatments. No significant differences were observed for pH, citric acid concentration and N-OPA content in both years. In 2016 the defoliation treatments showed lower values of N-OPA, while in 2017 OHS showed the lowest value. In both years the berry weight was significantly reduced by OHS. In 2016 the lower berry weight became apparent after véraison (Figure 4E), whereas in 2017 at véraison the berry weight of OHS was already significantly lower compared to the other treatments (Figure 4F).

Figure 4. Development of TSS (●) and sum (▲) of tartaric acid (TA) and malic acid (MA) for 2016 (A) and 2017 (B), of glucose/fructose ratio (●) and TA/MA ratio (▲) for 2016 (C) and 2017 (D) and of berry weight in 2016 (E) and 2017 (F); means (n=4) ± SD.

Save View full size Expand inline Collapse inline

The response of flavanols, hydroxycinnamic acids and flavonols, summarized as phenols, to light and heat differed dramatically within the treatments and with vintage (Table 5). The concentration of flavanols (procyandin B1, catechin, epicatechin) was always higher for the defoliation treatments in 2016. However, within this group no significant difference was observed between DEF_E and DEF_L. In 2017, there was no significant difference between the treatments for the same group of flavanols and the single compounds.

The concentration of hydroxycinnamic acids was the lowest in DEF_L and the highest in CON in 2016 and 2017. Only in 2016 were the results significant. In both years, CON had the highest concentration of caftaric acid and DEF_L had the highest concentration of fertaric acid. Comparing the years, there was no clear influence of the defoliations or the OHS-treatment on hydroxycinnamic acids, as a group or single compounds, except for fertaric acid, where the results were significant in both years.

The flavonol concentration was influenced most by the treatments. In both years, berries from the defoliation treatments had the highest concentration of flavonols and DEF_E always showed higher concentrations than DEF_L. In both years, que-3-rutinoside and que-3-glucuronide were highest for DEF_E. For DEF_L, the que-3-galactoside, que-3-glucoside, que-3-xyloside and que-3-rhamnoside were amongst the highest. The concentration of a single flavonol in a defoliation treatment compared to CON was up to five times higher. In 2016, for example, the mean concentration of que-3-rhamnoside in berry skin was 34 µg/g in CON and 147 µg/g in DEF_L.

In 2016, the highest concentration of phenols was found in DEF_E, followed by DEF_L, CON and OHS. OHS showed higher concentration of flavonols but lower concentration of hydroxycinnamic acids compared to CON. All three groups of flavanols, hydroxycinnamic acids and flavonols summarized as phenols, showed significant differences. In 2017 only the flavonol group showed significant differences across the treatments. Regarding the sum of phenols, DEF_E had the highest concentration again followed by DEF_L. Although not significant, OHS showed higher concentrations of phenols compared to CON which was caused by higher amounts of flavonols. The sum of phenols of CON was approximately 400 µg/g higher in berry skin in 2016 compared to 2017. For both defoliations and OHS, a rise of approximately 700 µg/g phenols in berry skin was observed.

The development of the phenols can be found in Figure 5. At the beginning of aroma development in 2016 (Figure 5A), the concentrations of DEF_E differed strongly compared to the other treatments, mainly caused by almost three times higher flavonol concentrations (DEF_E: 1145 µg/g, CON: 468 µg/g, DEF_L 485 µg/g and OHS 417 µg/g). DEF_L was initiated one week (71 DAF) before sampling (78 DAF) and behaved consistent with CON at this point in time. However, DEF_L resulted in a strong formation of phenols, especially of flavonols during aroma development at the end of the season and did not differ from DEF_E by harvest. In 2017, and prior to the treatment, both defoliation trials showed the same concentration of phenols. The treatment DEF_L was initiated two weeks (71 DAF) before the first phenol sampling (86 DAF). The concentration of phenols increased until the second sampling date and decreased close to harvest.

Figure 5. Changes in phenol concentration over time, from the beginning of aroma development to harvest in 2016 (A) and 2017 (B); means (n=4) ± SD.

Save View full size Expand inline Collapse inline

Figure 6 shows a principal component analysis (PCA) of phenols and maturity data from the years 2016 and 2017. For 2017, the PCA uses the data from Table 4 and Table 5 with the phenol levels and maturity determined at 106 and 109 DAF respectively. For 2016, where there was an 8-day difference between the sampling for phenols at 104 DAF and harvest at 112 DAF, maturity data (not shown) from 104 DAF has been used instead for the PCA. At this point in time the concentrations of sugars, malic acid and N-OPA were lower; tartaric acid and titratable acidity (TA) were higher compared to the concentrations of 2017. PC1 was mostly associated with variation between the vintages, with some separation of treatment effects. PC2 discriminated between treatments within season. Besides the influence of the vintage, the CON treatment is separated from the DEF treatment by PC2; OHS is intermediate.

In 2017 the phenol content was twice to that in 2016 which can be seen in the loadings plot where the phenols are oriented on the right-hand side of PC1. The fertaric acid, epicatechin and coutaric acid are found to be negatively correlated to the rest of the phenols (vintage effect). In both vintages the treatments affect the concentration of the phenols. The tartaric acid and the total acidity are higher in 2016 samples than in 2017. The malic acid and the tartaric acid are negatively correlated as could also be seen in loadings plot.

Figure 6. Scores and loadings plot of principal component analysis (PCA) of the last sampling date for phenols and corresponding maturity data in 2016 (104 DAF) and 2017 (106 DAF); TA= titratable acid, GRP= grape reaction product, p-CGT= p-coumaroylglycosyltartrate, Que= quercetin.

Save View full size Expand inline Collapse inline

Discussion

This study investigated the impact of elevated temperatures on Riesling berry composition using an open heating system (OHS) or various defoliation treatments to increase bunch exposure to solar radiation. Whilst defoliation affects mainly the surface temperature of the exposed berries, the peak temperature of individual berries may vary dramatically in absolute maximum values and its duration (Smart and Sinclair, 1976). In contrast, OHS increases the air temperature through the bunch regardless of the sun exposure. The increase in air temperature introduced through the OHS was around 5 °C at the opening and up to 2 °C at the edge of the bunch zone. Sadras and Soar (2009) achieved a maximum effect of 2-4 °C at the bunch height and obtained similar results to those achieved in this study. Their row distance (3.0 m) was wider compared to our row distance (2.1 m) with reduced shading effects. However, the efficiency of such passive heating systems depends on the frequency of radiation days and sunshine hours which occurred more frequently at the experimental site located in Australia.

Characterization of canopy structure and light conditions inside the canopy by point quadrat and light sensitive films showed significant differences between control (CON) and defoliations (DEFs). The use of light sensitive films allowed quantification of the cumulative solar radiation inside the canopy over a short time period. The films were placed in the middle of the canopy and had the lowest sun exposure. It has been shown recently, that the light sensitive films are not overly sensitive to temperature (Bontempo et al., 2018). A further advantage of the films is that they provide cumulative measurements of solar radiation over a measured time period. To relate point quadrat readings to quantified measurements using light sensitive films, it was concluded that the surface area of the films is rather small (20 mm x 35 mm), and that a film might be shaded by a bunch or a leaf. This may explain the different results for the treatments in 2017, where the first point quadrat result showed significant differences between the treatments whilst the measurement of solar radiation by the films did not show a significant difference over the same time period. This variation shows the need for different measurements to characterize the canopy structure and solar radiation in the canopy.

Temperature influences the phenological development of the vines. Higher temperatures lead to an earlier budburst, flowering and ripening. Temperature also has an impact on berry weight. Hale and Buttrose (1974) showed temperatures above 30 °C reduced the berry size. In 2017 the temperature effect of the OHS had a strong influence as the berry weight was 17% lower than the other treatments over the season. In 2016, a decrease in the berry weight of 5% was observed at the end of the season. This less intense reduction in berry weight with OHS compared to CON observed in 2016 may be caused by the high precipitation and low sun radiation at the beginning of ripening. After véraison, effects of temperature on berry size are less noticeable (Keller, 2010). Bonada et al. (2015) used an open top chamber with and without irrigation and observed a decrease of 8% in berry weight for heated berries without irrigation, but little effect (-2%) on berry weight for heated berries with irrigation. Furthermore, higher temperatures are likely to increase to total soluble solids (TSS) and produce lower titratable acidity (TA) in Riesling (Bock et al., 2011; Duchêne and Schneider, 2005; Urhausen et al., 2011). For Riesling, a rise of approximately 0.6 °Brix per decade was observed between 1949-2010 in Franconia, Germany (Bock et al., 2011). However, TSS accumulation can also be decreased by temperatures above 30 °C (Kriedemann and Smart, 1971). Greer and Weston (2010) showed in their studies also, the time period (flowering, fruit set, véraison, mid ripening) of hot weather conditions (40/25 °C day/night temperatures over 4 days) had an influence on the rate of sugar accumulation. In 2016, budburst was at the same date as for the 30-year average (30-AVG, 1981-2010). In 2017, the early warm period caused a 14 days earlier budburst compared to the 30-AVG. At the first sampling date in 2017 (72 days after flowering (DAF)), TSS was 1.5 °Brix higher in OHS compared to CON. At the last sampling date in 2017 (109 DAF), OHS had approximately 1 °Brix less compared to CON. At the beginning of the season the ripening of OHS was advanced compared to the other treatments. This result confirms the results of other studies, showing an earlier ripening caused by warmer temperatures. During the season, high temperatures from the OHS may have led to a decline in sugar accumulation and resulted in a lower TSS content. In 2016, at the beginning of sampling (71 DAF), no difference between CON and OHS was recorded while at the end of the season (112 DAF) OHS had 1 °Brix less compared to CON, as found in 2017. Defoliation treatments had no significant influence on TSS, which was also observed for Riesling by Friedel et al. (2015). In both years the glucose to fructose ratio significantly decreased for OHS compared to the other treatments. Kliewer (1967) showed a decrease in the glucose to fructose ratio in vintages with higher temperatures.

It has been noted in many other studies that the acidity content decreases under warmer temperatures due to malic acid respiration (Buttrose et al., 1971; Lakso and Kliewer, 1975; Ruffner et al., 1976), whereas the concentration of tartaric acid is unaffected (Kliewer, 1965; Rienth et al., 2016). At the beginning of véraison in 2017, the mean malic acid concentration was 14.2 g/L in OHS compared to CON (19.4 g/L) a reduction of approximately 27%, while prior to harvest it was approximately 42% less in OHS (3.4 g/L) compared to CON (5.8 g/L). In 2016 the concentration of malic acid was approximately 3% less for OHS (19.9 g/L) compared to CON (20.6 g/L), for first sampling date and 18% less for OHS (2.8 g/L) compared to CON (3.4 g/L) at the last sampling date. The ratio of tartaric acid to malic acid is significantly influenced by temperature effects in both years. Tartaric acid is synthesized during early stages of berry development. In both years, DEF_L had the lowest concentration of tartaric acid, even though DEF_L and CON had the same conditions during early stages until véraison. In both years CON showed the highest values of shikimic acid at the last sampling date. In 2016, DEF_L had the lowest concentration of shikimic acid. This might be due to the intense sun radiation shortly after the defoliation. In 2017, OHS showed the lowest concentration of shikimic acid followed by DEF_L. In both years the results were significantly different. Even though the concentration of malic acid and consequently the TA were significantly decreased by OHS, no effect on the pH was observed. The yeast available nitrogen content expressed as N-OPA, was lower in 2016 for the DEFs compared to CON and OHS. Our results in 2016 point towards the results of Friedel et al. (2015), who showed that exposed berries had lower N-OPA concentrations compared to shaded berries. In 2017, the DEFs had similar concentrations compared to CON, whereas OHS showed the lowest concentration.

The sum of flavanols was significantly increased for DEFs compared to CON and OHS with a sunny period during aroma development in 2016. In 2017, no differences between treatments were observed during low levels of solar radiation in the aroma development period. Friedel et al. (2015) did not observe an influence of shading or leaf removal on flavanol concentrations, with only a slight increase in catechin. Koyama et al. (2012) showed a light-induced increase in flavanol accumulation for higher porosity treatments when applied at flowering. At the first sampling date in 2016, higher flavanol levels occurred for DEF_E while at the last sampling date, DEF_E and DEF_L showed similar elevated levels compared to CON and OHS. In contrast to Friedel et al. (2015) an influence of defoliation was observed when high solar radiation levels occurred during aroma development. In contrast to Koyama et al. (2012), the DEF_L treatment showed higher flavanol concentrations. The influence of temperature on flavanol biosynthesis is not well understood (Gouot et al., 2019). The in-vitro studies of Ayenew et al. (2015) and Degu et al. (2016) found effects of high temperatures on flavanols in red berry skin and showed conflicting results, for epicatechin for example. Ayenew et al. (2015) reported a decrease in epicatechin and procyanidin, whilst Degu et al. (2016) showed an increase in epicatechin with higher temperatures.

The influence of the treatments on hydroxycinnamic acid concentration differed over time. In both years, DEF_L showed low concentrations of caftaric acid, coutaric acid and cumarolyglucose and high concentrations of caffeic acid and fertaric acid compared to CON. DEF_L and OHS had mostly similar concentrations. DEF_E showed similar concentrations compared to CON. A similar treatment effect comparing DEF_L and CON was observed by Friedel et al. (2015). Limited data is available for hydroxycinnamic acid in Vitis vinifera, with most studies investigating the influence of light.

The greatest influence of light was observed for flavonols through different timing of defoliation treatments. In 2016 at the first sampling date (78 DAF), DEF_L had the same concentrations of flavonols as CON and OHS, whilst DEF_E showed almost three times higher concentrations. During the aroma development and at the last sampling date (104 DAF), DEF_L showed almost the same flavonol concentrations as DEF_E, whereas CON and OHS remained almost unchanged from the earlier sampling dates. The flavonol concentration of DEF_E only increased by 10%, whereas with DEF_L the increase was 250%. In 2017, at the last sampling date (106 DAF) the increase of flavonols for OHS was twice as high compared to CON. DEF_E showed a 30% increase in flavonols and DEF_L a 20% increase. Both DEFs showed more than two times higher concentrations of flavonols compared to CON. From the implementation of DEF_E, until the first sampling date for phenols, 430 SSH were recorded in 57 days in 2016. The berries of DEF_L were exposed for 64 SSH in 6 days until the first sampling. In 2017 DEF_E had 424 SSH over 64 days and DEF_L 103 SSH over 12 days from treatment initiation until the first sampling. This may explain the different levels of flavonols for the early and late DEF at the first sampling date. In 2017, there was more time from treatment initiation until the first sampling point for DEF_L which was influenced by more SSH giving similar flavonol levels to DEF_E. In 2016, the time was too short to reach higher levels compared to CON at the first sampling date. From the first to the last sampling date 189 SSH (over 26 days) in 2016 and 94 SSH (over 23 days) in 2017 were recorded. The high solar radiation in 2016 produced a rapid increase in the concentration of flavonols for DEF_L. Mori et al. (2005) investigated the effect of elevated night temperatures (30 °C vs. 15 °C) and constant 30 °C day temperature on red grape berry flavonols. Similar to the findings of Spayd et al. (2002), Mori et al. (2005) also found no effect of temperature treatments on flavonols.

The PCA showed separation of vintages on PC1, and separation of the treatments of PC2. The impact of vintage was stronger than the treatment effect, which may reflect the contrasting weather conditions of the two seasons. However, the distribution of the treatments in both years was very similar, indicating similar relative responses of the defoliation and heating treatments compared to the control. Flavonols had strong loadings on PC1 and PC2, indicating their strong responsiveness to weather conditions and treatments. Phenols are produced via metabolic pathways which are driven by enzymes and influenced by temperature, light and plant water status (Jackson and Lombard, 1993). Consequently, a direct relationship between the concentration of phenols and the influence of defoliation and heat treatments cannot be determined. In summary, the phenol concentration was mainly impacted by defoliation. The impact of temperature on phenols was small or non-existent, depending on the timing of heat effects on the bunches.

Higher phenol content in berry skin leads to an astringent taste of the wine. For Riesling its fruity character and a balance between sweetness and acidity is desired. Trying to mimic future scenarios of increasing temperature through OHS, it was found that malic acid was reduced, smaller berries were produced, and the glucose/fructose ratio changed with potential impacts on wine style. Furthermore, a high degree of bunch exposure strongly increased phenol content with potential impacts on astringency of the taste and, to a lesser extent, on sugar and acidity balance.

Defoliation practices have been considered as a powerful tool to improve berry health and quality. However, under higher temperature, these practices in Riesling will need to be considered more carefully in relation to timing and intensity in order to maintain a desired wine profile.

Acknowledgements

The authors would like to thank the team of the Department of the General and Organic Viticulture and especially Annette Rheinberger, Heike Semmler and Jeannette Zacharias for their technical support. We also thank Jason Smith for assistance with the manuscript revisions; we acknowledge the financial support for the research provided by the national ministries of Aquitaine, South Australia and Hesse for funding the research alliance BAG (Bordeaux, Adelaide, Geisenheim).

References

• Arnold R.A., Noble AC., and Singleton V.L., 1980. Bitterness and astringency of phenolic fractions in wine. Journal of Agricultural and Food Chemistry, 28, 675–678. doi:10.1021/jf60229a026
• Ayenew B., Degu A., Manela N., Perl A., Shamir M.O. and Fait A., 2015. Metabolite profiling and transcript analysis reveal specificities in the response of a berry derived cell culture to abiotic stresses. Frontiers in Plant Science, 6. doi:10.3389/fpls.2015.00728
• Batista L., Monteiro S., Loureiro V.B., Teixeira AR. and Ferreira R.B., 2010. Protein haze formation in wines revisited. The stabilising effect of organic acids. Food Chemistry, 122(4), 1067–1075. doi:10.1016/j.foodchem.2010.03.076
• Blancquaert E.H., Oberholster A., Ricardo-da-Silva J.M. and Deloire A.J., 2019. Effects of Abiotic Factors on Phenolic Compounds in the Grape Berry – A Review. South African Journal of Enology and Viticulture, 40(1). doi:10.21548/40-1-3060
• Bock A., Sparks T., Estrella N. and Menzel A., 2011. Changes in the phenology and composition of wine from Franconia, Germany. Climate Research, 50(1), 69–81. doi:10.3354/cr01048
• Bonada M., Jeffery D.W., Petrie P.R., Moran M. A. and Sadras V.O., 2015. Impact of elevated temperature and water deficit on the chemical and sensory profiles of Barossa Shiraz grapes and wines. Australian Journal of Grape and Wine Research, 21(2), 240–253. doi:10.1111/ajgw.12142
• Bonada M. and Sadras V.O., 2015. Review: Critical appraisal of methods to investigate the effect of temperature on grapevine berry composition. Australian Journal of Grape and Wine Research, 21(1), 1–17. doi:10.1111/ajgw.12102
• Bontempo A., Smith J., Brandt M., Peterlunger E. and Stoll M., 2018. Evaluation of photosensitive films for light measurements in the fruiting zone of grapevine canopies. Vitis, 57, 159–165. doi:10.5073/VITIS.2018.57.159-165
• Buttrose M.S., Hale C.R. and Kliewer W.M., 1971. Effect of temperature on the composition of 'Cabernet Sauvignon' berries. American Journal of Enology and Viticulture, 22(2), 71–75.
• Coombe B.G., 1995. Adoption of a system for identifying grapevine growth stages. Australian Journal of Grape and Wine Research, 1(2), 104–110. doi:10.1111/j.1755-0238.1995.tb00086.x
• Cubasch U., Wuebbles D., Chen D., Facchini M.C., Frame D., Mahowald N., and Winther J.-G., 2013. The Physical Science Basis. Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change, [Stocker, T.F., D. Qin, G.-K. Plattner, M. Tignor, S.K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex and P.M. Midgley (eds.)].
• Degu A., Ayenew B., Cramer G.R. and Fait A., 2016. Polyphenolic responses of grapevine berries to light, temperature, oxidative stress, abscisic acid and jasmonic acid show specific developmental-dependent degrees of metabolic resilience to perturbation. Food Chemistry, 212, 828–836. doi:10.1016/j.foodchem.2016.05.164
• Duchêne E. and Schneider C., 2005. Grapevine and climatic changes: A glance at the situation in Alsace. Agronomy for Sustainable Development, 25(1), 93–99. doi:10.1051/agro:2004057
• Dukes B.C. and Butzke C.E., 1998. Rapid determination of primary amino acids in grape juice using an o-phthaldialdehyde/N-acetyl-L-cysteine spectrophotometric assay. American Journal of Enology and Viticulture, 49(2), 125–134.
• Ferreira R. B., Piçarra-Pereira M.A., Monteiro S., Loureiro V.B. and Teixeira A.R., 2001. The wine proteins. Food Science and Technology, 12, 230–239. doi:10.1016/S0924-2244(01)00080-2
• Friedel M., Stoll M., Patz C.D., Will F. and Dietrich H., 2015. Impact of light exposure on fruit composition of white 'Riesling' grape berries (Vitis vinifera L.). Vitis, 54, 107–116.
• Gouot J.C., Smith J.P., Holzapfel B.P., Walker A.R. and Barril C., 2019. Grape berry flavonoids: a review of their biochemical responses to high and extreme high temperatures. Journal of Experimental Botany, 70(2), 397–423. doi:10.1093/jxb/ery392
• Greer D.H. and Weston C., 2010. Heat stress affects flowering, berry growth, sugar accumulation and photosynthesis of Vitis vinifera cv. Semillon grapevines grown in a controlled environment. Functional Plant Biology, 37(3), 206–214. doi:10.1071/FP09209
• Hale C.R. and Buttrose M.S., 1974. Effect of temperature on ontogeny of berries of Vitis vinifera L. cv. Cabernet Sauvignon. Journal of the American Society for Horticultural Science, 99(5), 390–394.
• Intergovernmental Panel on Climate Change., 2014. Climate Change 2013 - the physical basis: Working Group I Contribution to the Fifth Assesment Report of the Intergoverment Panel on Climate Change. Cambridge: Cambrige Universtiy Press.
• Jackson D.I. and Lombard P., 1993. Environmental and Management Practices Affecting Grape Composition and Wine Quality - A Review. American Journal of Enology and Viticulture, 44(4), 409–430.
• Keller M., 2010. The science of grapevines: Anatomy and physiology. New York: Academic Press.
• Kliewer W.M., 1965. Changes in the concentration of malates, tartrates, and total free acids in flowers and berries of Vitis vinifera. American Journal of Enology and Viticulture, 16(2), 92-100.
• Kliewer W.M., 1967. The glucose-fructose ratio of Vitis vinifera grapes. American Journal of Enology and Viticulture, 18, 33–47.
• Knoll C., Fritsch S., Schnell S., Grossmann M., Krieger-Weber S., Du Toit M. and Rauhut D., 2012. Impact of different malolactic fermentation inoculation scenarios on Riesling wine aroma. World Journal of Microbiology and Biotechnology, 28(3), 1143–1153. doi:10.1007/s11274-011-0917-x
• Komm B.L. and Moyer M.M., 2015. Effect of Early Fruit-Zone Leaf Removal on Canopy Development and Fruit Quality in Riesling and Sauvignon blanc. American Journal of Enology and Viticulture, 66(4), 424–434. doi:10.5344/ajev.2015.15007
• Koyama K., Ikeda H., Poudel P. R. and Goto-Yamamoto N., 2012. Light quality affects flavonoid biosynthesis in young berries of Cabernet Sauvignon grape. Phytochemistry, 78, 54–64. doi:10.1016/j.phytochem.2012.02.026
• Kriedemann P. and Smart R., 1971. Effect of irradiance, temperature and leaf water potential on photosynthesis of vine leaves. Photosynthetica, 5, 6–15.
• Lakso A.N. and Kliewer W.M., 1975. The Influence of Temperature on Malic Acid Metabolism in Grape Berries. Plant Physiology, 56, 370–372. doi:10.1104/pp.56.3.370
• Mira de Orduña R., 2010. Climate change associated effects on grape and wine quality and production. Food Research International, 43(7), 1844–1855. doi:10.1016/j.foodres.2010.05.001
• Mori K., Sugaya S. and Gemma H., 2005. Decreased anthocyanin biosynthesis in grape berries grown under elevated night temperature condition. Scientia Horticulturae, 105(3), 319–330. doi:10.1016/j.scienta.2005.01.032
• Rienth M., Torregrosa L., Sarah G., Ardisson M., Brillouet J.-M. and Romieu C., 2016. Temperature desynchronizes sugar and organic acid metabolism in ripening grapevine fruits and remodels their transcriptome. BMC Plant Biology, 16(164. doi:10.1186/s12870-016-0850-0
• Ruffner H.P., Hawker J.S. and Hale C.R., 1976. Temperature and enzymic control of malate metabolism in berries of Vitis vinifera. Phytochemistry, 15, 1877–1880. doi:10.1016/S0031-9422(00)88835-4
• Sadras V.O. and Soar C.J., 2009. Shiraz vines maintain yield in response to a 2–4°C increase in maximum temperature using an open-top heating system at key phenostages. European Journal of Agronomy, 31(4), 250–258. doi:10.1016/j.eja.2009.09.004
• Singleton V.L., 1987. Oxygen with phenols and related reactions in must, wines, and model systems: observations and pratical implications. American Journal of Enology and Viticulture, 38, 69–77.
• Singleton V.L., Sieberhagen H.A., Wet P. de and van Wyk C.J., 1975. Composition and sensory qualities of wines prepared from white grapes by fermentation with and without grape solids. American Journal of Enology and Viticulture, 26, 62–69.
• Smart R. and Robinson M., 1991. Sunlight into wine: a handbook for winegrape canopy management. Adelaide: Winetitles.
• Smart R. and Sinclair T.R., 1976. Solar heating of grape berries and other spherical fruits. Agricultural Meteorology, 17(4), 241–259. doi:10.1016/0002-1571(76)90029-7
• Spayd S.E., Tarara J.M., Mee D.L. and Ferguson J.C., 2002. Separation of Sunlight and Temperature Effects on the Composition of Vitis vinifera cv. Merlot Berries. American Journal of Enology and Viticulture, 53(3), 171–182.
• Urhausen S., Brienen S., Kapala A. and Simmer C., 2011. Climatic conditions and their impact on viticulture in the Upper Moselle region. Climatic Change, 109, 349–373. doi:10.1007/s10584-011-0059-z
• Vršič S. and Vodovnik T., 2012. Reactions of grape varieties to climate changes in North Slovenia. Plant Soil Environment, 58(1), 34–41. doi:10.17221/352/2011-PSE