The development of new technologies in agriculture is enabling major improvements in the quality of agricultural products, such as the technologies for managing the various vegetative vigors.

By Leonardo D’Intino

Master Thesis in Agricultural Sciences and Technologies – Enology and Viticulture (October 2023) – Catholic University of the Sacred Heart – Thesis supervisor Mario Gabrielli leonardo.dintino1@unicatt.it; mario.gabrielli@unicatt.it

According to scientific literature available, the vegetative vigour (growth rate of a vegetative system) affects the quality of agricultural products. Even in the grapevine, the vigour leads to differences in the composition of the grapes; for example, vigorous vines tend to produce grapes with higher sugar content, more pronounced acid profiles and less accumulation of secondary metabolites (polyphenols and aromas). On the other hand, less vigorous vines produce grapes that are richer in sugar and secondary metabolites and less acidic. The final quality of the wines is defined not only by the alcohol content (% v/v) and the total acidity (g/l) but, above all, by secondary metabolites such as polyphenols (mg/L) and aromas (<mg/L).

In this respect, not all grapes have the same aromatic endowment. Most varieties of Vitis vinifera have a simple taste, i.e. low concentrations of primary aromas, produced directly from grapes such as terpenes (unlike esters and alcohols produced by fermentation). Terpenes are a class of metabolites consisting of the assembly of several isoprene units; depending on the number of units different subclasses are obtained, of which monoterpenes (two units) are the most common. Terpenes are essential compounds for all aromatic wines, that is, wines produced from aromatic grapes in which the concentrations of the different aromas exceed the threshold of olfactory perception obtaining wines that are very fruity, floral and with the development of balsamic.

The wines produced from certain Malvasia vines are an example of this, such as Malvasia di Candia Aromatica, with which the wines covered by the study were produced. It is known in the literature that the plant’s vigour strongly influences the accumulation of secondary metabolites; therefore, you can expect very different wines from zones with different vigour. With this in mind, dividing wine productions according to their vigour, and their separate processing and marketing can be a powerful tool for differentiating farm productions (range extension), increasing product quality (more homogeneous mass winemaking), and a powerful marketing driver from the vineyard.

The different grapes and wines have no different quality, that is, it is wrong to speak of high and low quality. In fact, wines produced from the more vigorous areas are generally fresher (more acidic), less alcoholic, and intended for faster consumption, while wines from the less vigorous areas are rich and concentrated, more alcoholic and require longer ageing. The production of two different wines from the same vineyards allows the producer to obtain a first range of quick-to-market products, with which to recoup the investments of the harvest, and a second range of products, with longer production times that will be marketed later.

Aim

The aims of this thesis were many: on the one hand, the evaluation of wines produced from mechanized and selective harvesting based on vigour, and on the other hand the evolution of the wines themselves over time and based on storage temperatures. For the first part, both the basic parameters and the aromatic profile were assessed, while the evaluation of wines during shelf-life focused on the evolution of the aromatic profile alone.

The particular focus given to the aromatic profile is since few variations were expected on the basic parameters, and that the aromatic profile of wines obtained from varieties such as Malvasia di Candia aromatica is probably the main qualitative parameter. The final aim of the thesis was therefore to enable producers to know both the effects on quality of the selective harvest and to know the best time for marketing the various wines produced.

Materials and methods

Vineyard

The vineyard used for this test is in the “Villarosa” estate (Bacedasco, Vernasca, PC), and is part of the PSR project RIPrESO (Prof. Gatti Matteo: matteo.gatti@unicatt.it). The vineyard was planted in 2004 with Malvasia di Candia Aromatica. Within the vineyard, an area of 15 rows has been identified, which has been characterized according to vigour through a MECS-VINE sensor. This sensor is applied directly to the tractor and records several parameters while it moves between the rows, ultimately being able to determine a vigour index (CI, canopy index) in combination with GPS coordinates.

The study area was then divided into 3 sub-areas according to vigour: High vigour (HV), low vigour (LV) and control (C, containing both vigorous and non-vigorous areas). For each zone of vigour, 3 blocks were identified, for a total of 9 parcels. The obtained vigour map was loaded on a mechanical grape harvester (New Holland Braud 9080L) to divide the grapes coming from the different areas (HV, LV, C). For each parcel, 30 kg of grapes were taken directly from the grape harvester, immediately chilled and transferred to the experimental cellar of the Catholic University of the Sacred Heart, Piacenza, where they were vinified.

Microvinifications

The microvinifications followed the following protocol: each mass of grapes was pressed with a hydropress and the must obtained was placed in small 30L stainless steel vats, protected from oxygen to preserve the terpenes, sulphitated and inoculated. At the end of fermentation and after two pourings, the wines were bottled in dark glass with a crown cap and analysed to determine their basic oenological parameters and aromatic profile before starting the shelf-life study.

Shelf-life parameters

Shelf-life was selected to assess both the effect of temperature and time. The different temperatures were 5°C (fridge temperature), 15°C (cellar storage) and 25°C (domestic storage). For each temperature, sampling was carried out every 3 months (0, 90 and 180 days); then for each period, 27 bottles were collected (9 bottles for each temperature, one bottle for each parcel for each temperature). At T0 the basic parameters and aromas were analysed, at T1 and T2 only aromas.

Methods

The basic parameters were analyzed according to official methods (OIV, 2024). The volatile compounds were analyzed using HS-SPME-GC×GC-MS, a two-dimensional gas chromatograph (GC×GC), with fiber headspace sampler (HS-SPME) and identification of compounds by mass spectrometry (MS). The use of a GC×CG, with two columns of opposite polarity, allows excellent chromatographic separation, greatly reducing any possible co-elution.  Each sample was prepared by diluting 1:4 with water and 1 ml of wine and adding 2 g of NaCl and 100 ml of internal standard (2-octanol) in 25 ml vials.

Extraction of volatile compounds from the vial headspace was performed with an AOCTM-6000 autosampler (Shimadzu Co., Kyoto, Japan) with a stirrer and SPME fiber. The GC×GC system consisted of a NexisGC-2030 coupled with a TQ8040NX mass detector. The chromatographic run was performed on two columns, with the first SLB-5ms non-polar column coupled with the second SupelcoWAX polar column. The mass spectrometer has been set to a mass range of 40-360 m/z. The resulting chromatograms were integrated, and each peak was identified by comparison on the NIST20 library. Only identifications with similarities greater than 85% were considered. Determination of the concentration of the individual compounds was obtained through the area ratio to the internal standard.

Statistical analysis

An ANOVA (variance analysis) was carried out on the basic oenological parameters and aromas at T0, while a statistical model of PLS-DA (partial least square-discriminant analysis) coupled with the fold change analysis (FC) was used to determine the time and temperature effect. The use of these statistical models served respectively to reduce the complexity of the dataset by extracting only the most significant compounds (PLS-DA with VIP compound extraction) and to describe the trends of these compounds (FC: Increase/decrease in concentration).

Results

Basic parameters

As shown in Table 1, at the time of harvest the most significant figure was the difference in sugar concentrations (approx. 10 g/L difference between LV and HV), while minor differences were found in the acid profile. The data on must found consistency with that of wine, where ethanol was the most significant parameter among the various theses and was about 0.6 v/v higher in LV than HV. On the contrary, the total acidity values were higher in HV than in LV with about a 0.5 g/L difference; the pH data were in line with those of titratable acidity. As regards the profiles of individual acids, the only significant difference was citric acid, with higher values in HV than LV.

The data reported are consistent with what is already known in the literature, namely that different vines lead to different must compositions: Higher concentrations of sugar in the low vigour and higher concentrations of acidity in the high vigour. The ethanol figure, in addition to being consistent with the sugar content of the must, was therefore consistent with the higher alcoholic strength of wines produced from grapes of less vigorous vines.

T0 aromas

A total of 112 aromas, of which 63 terpenes (aromas of Fruity, floral, herbaceous and balsamic), 34 esters (fruity, floral and alcoholic), 5 norisoprenoids (balsamic), 5 alcohols (floral and green) and 5 others (aldehydes, ketones and naphthalenes) were found on T0 wines.

As shown in Figure 1, esters were the most present family in terms of concentration (approx. 70%), followed by alcohols (20%), terpenes (10%) and norisoprenoids (0.25%). This composition confirms an aromatic profile with fermentative aromas and a strong terpenic kit with some norisoprenoids. Although alcohols and esters were present in higher contractions, their aromatic impact is less than that of terpenes, which are characterized by very low thresholds of perception (mg/L).

Figure 1: T0 aromas, grouped by family (esters, alcohols, terpenes and norisoprenoids). Different letters indicate significantly different groups according to the Duncan test (P < 0.05).

The family that had the most significantly different compounds was that of terpenes, with 19 compounds, as shown in Table 2, where only the different terpenes among the various theses are listed. Non-different terpenes and the other aromatic families were deliberately omitted. The most relevant of these were linalool and geraniol, with each of these LCs showing higher concentrations than LV and C, a trend similar to that of the other compounds reported. The data presented are consistent with what is already known in the literature: Less vigorous vines tend to accumulate more terpenes than the more vigorous vines.

It can be concluded that the extraction and analysis technique has been able to well define the aromatic profiles of the wines and to identify a large number of aromas. The low concentration values for alcohols (reported in the literature in the order of hundreds of mg/L) are due to the low affinity of fiber toward this family of compounds, a necessary compromise to increase selectivity toward terpenes and norisoprenoids.

 Effect of Temperature

The PLS-DA model used allowed the identification of 48 out of 122 VIP compounds, which better describe the effect of storage temperature on aromas. Among these were 23 terpenes, 12 of which increased and 11 decreased, as described by FC. As reported in the literature, the decrease of the most characteristic terpenes of young aromatic wines (fruity/floral/herbaceous notes) and the increase of the derived forms – oxidized and esterified (balsamic notes) – were expected. For example, linalool decreased as its derived forms increased. For aromas belonging to the other aromatic classes, esters decreased as the storage temperature increased, while norisoprenoids increased, following the literature available on the subject.

Figure 2: This graph describes the temperature effect (PLS-DA). In particular, component 1 (describing 41 % of the variability between samples) was able to discriminate the effect of temperature. It can also be seen that the 5th and 15th samples were closer to each other (and therefore more similar) than those at 25°C.

Effect of time

Here, too, the PLS-DA model has allowed us to highlight the 44 VIP compounds to discriminate the effect of storage time. 19 were terpenes, of which 11 decreased over time while only 8 increased. Most terpenes that had the most impact on the aromatic profile showed a reduction in concentration over time, such as linalool (fruity and floral) while others such as β-myrcene (with more spicy and balsamic hints) increased. These changes confirm that ageing tends to increase the concentration of spicy compounds compared to the fruity and floral notes of young wines. Unlike the temperature effect, more VIPs had a negative tendency following the available literature on the shelf-life of wine, the majority of VIPs were reduced and the aromatic intensity of the wine was reduced.

Figure 3: This graph describes the effect of time (PLS-DA). In particular, component 1 (describing 36 % of the variability between samples) was able to discriminate against the effect of the shelf-life. The greater distance between the blue cluster (T2) and the red and green cluster (T0 and T1) indicates that there were greater differences between T1 and T2 than between T1 and T0.

Effect of time and temperature

In addition to the above, an ANOVA was carried out at T2 of the wines stored at 15 °C. Lower differences compared to T0 were found: The sum of terpenes was significantly higher in LV but of the 122 aromas considered, only 9 were significantly different and, except for γ-terpinene, none with a significant aromatic impact. Therefore, the shelf-life reduced the differences between the theses, which meant that the wines were more aromatically similar.

Conclusions

As expected, both grapes and must revealed differences in sugar, and thus wines in ethanol, in contrast to other parameters such as acids that showed less statistical significance. As regards aromas, differences in terms of families and individual compounds were noted, especially regarding terpenes, which are the most relevant compounds for a young aromatic wine.

Just to mention the two of the most important terpenes in defining the aromatic profile of wines: Linalool and Geraniol, the different theses of vigour (HV and LV) showed significant differences, with a higher concentration in LV than HV, following what is already known in the literature. At the same time, some terpenes seem to be less affected by vigour. Several results on norisoprenoids showed a lesser response to vigour, none of the theses being significantly different. Finally, the statistical models adopted have been effective in discriminating against different conservation conditions and identifying trends