Trellis effects on wine composition (Part 2): Aroma

by | Aug 1, 2021 | Oenology research, Viticulture research, Winetech Technical

Yield is a primary objective in the industry and can be maximised by trellising systems. Besides yield per vine, synthesis and further evolution of chemical compounds may be altered by vine architecture, and this is the core subject for the current study.

The aim of this study was to evaluate the influence of six different trellis systems in South African conditions. Chenin blanc vines were trained on Santorini, Ballerina, Smart Dyson, T-frame, Lyre and Stok-by-paaltjie. Grapes from two consecutive vintages (2017 and 2018) were harvested at various dates corresponding to 22 – 23°Brix and wine was made with the same protocol. No wood contact was used. The wines spent three months on lees before bottling and six months in the bottle before chemical and sensorial analysis.

The classes of aroma compounds analysed in the wines were the ones most likely to be affected directly or indirectly by the vineyard practices, such as thiols and major volatiles, respectively. The wines were profiled by industry professionals and by an analytical panel at the Department of Viticulture and Oenology using Check-All-That-Apply (CATA) method. The profiles were evaluated to see if we could distinguish between the wines corresponding to the different trellising systems. Quality was also assessed by industry professionals for the same reason.

Chemical analyses

A total of 25 major volatiles and three thiols were determined in the wines. The aroma compounds were evaluated as individual compounds and as classes (esters, acetates, ethyl esters, alcohols, acids, total major volatiles, and thiols). The wines from horizontal dividing or open canopy type of trellis (Lyre and T-frame) had higher concentrations of thiols and major volatiles than the rest of the systems, although comparable in practical terms. Despite some significant differences in some of the individual compounds (for example 3MHA was the highest in Lyre wines in the second season), overall differences were not apparent for either vintage. In the multivariate analysis, some of the samples corresponding to different systems clustered together; this was as a result of the similarity or dissimilarity between the systems and not due to significant variation between repeats.

Aroma profiles

CATA is a rapid sensory profiling technique, which uses a list of attributes (in the form of words or phrases), from which trained or untrained panellists can select all the descriptors they consider appropriate to characterise each sample (Valentin et al., 2012). CATA aroma terms used in this study were selected from the South African Chenin blanc aroma wheel. The results were compiled first in a frequency of citation table and subsequently submitted to multivariate analysis.

For the first season (2017), there were few differences in terms of number and type of descriptors used for the wines: 25 – 30 for the industry professionals, while the analytical panel used all 40 terms in the list. The most cited attributes overall were ‘passion fruit’ (professionals) and ‘pineapple’ (analytical panel). ‘Passion fruit’ was also used by the analytical panel for all the wines; ‘guava’ and ‘grapefruit’ also appeared in the top five across all systems. Looking at the industry professionals’ list, ‘pineapple’ was frequently used across all systems, together with ‘lemon’, ‘melon’ and ‘peach’. This was already an indication that the wines might be quite similar across treatments.

The configuration resulting from the multivariate analyses demonstrated what was already suggested by the frequency data, all samples overlapping except for Lyre wines. These wines were described by the professionals with ‘passion fruit’ and ‘grapefruit’ (common attributes for all wines), but also ‘vanilla’, ‘baked bread’ and ‘stewed fruit’ (less common across the systems). The analytical panel described only the Lyre wines with ‘fynbos’ and ‘oak’.

In 2018, only the analytical panel tasted the wines. This time, 36 terms were used to describe the wines and again ‘pineapple’ was the most cited. Overall, three of the top five attributes were common to all treatments for both years – ‘pineapple’, ‘apple’ and ‘passion fruit’. However, for this vintage there were some differences resulting from the multivariate analysis configuration: Lyre and Santorini wines were associated with ‘floral’, ‘pear’ and ‘dusty’; T-frame wines had the most (overall) common attributes ‘orange’, ‘peach’, ‘apple’, ‘lemon’ and ‘passion fruit’; Ballerina wines were characterised by ‘marmalade’, ‘fynbos’ and ‘tobacco’; and Smart-Dyson wines were ‘spicy’, ‘caramel’, ‘papaya’ and ‘sweet associated’.

Quality evaluation

The industry professionals evaluated the wines only in 2017; they were instructed to use the 20-point scale and judge the colour out of 3, aroma out of 7, and taste out of 10. From a quality perspective, the aromas of all wines were judged to be similar. For aroma, only Santorini wines were statistically lower than the rest (average of 4.5 out of 7), while Lyre wines were evaluated as the highest (average of 5 out of 7). There were no differences for colour and all wines scored close to 3. Santorini wines were once more evaluated as the lowest for taste; all the other wines were scored similar to each other. Overall, Santorini wines were the lowest scoring (12 out of 20, statistically lower than the rest). Lyre and T-frame scored the highest (approximately 14 out of 20).

Take home message

Even though a lack of differences between systems can be considered a negative result, in the case of the current study it is a positive one. It means that the trellising systems did not affect the wines extensively. In this case, the choice can be a system that results in increased yield without compromising the aroma. Chemically, the wines were not identical, but the overall composition did not differ significantly. A similar trend was observed for the aroma data of 2017 and 2018; all wines were described by fruity notes which relate to esters and thiols. The floral notes mentioned could be associated with other aromatic chemical compounds (terpenes), not analysed in the study.

The aroma quality was also not compromised. The results for the other two aspects of quality are somewhat different; colour, taste and mouthfeel are mainly related to non-volatile compounds, and these results will be presented in the Part 3 of this series.


Are wine aroma characteristics influenced by different trellising systems?

Despite the importance of trellising systems in viticulture, little work has been done on the specific effect of various trellising systems on the wine’s chemical and sensorial characteristics. To answer this, it is of interest to compare wines of various trellis systems from a chemical and sensorial point of view, under controlled conditions.

– For more information, contact Astrid Buica at


Article Archives

Search for more articles

Generic selectors
Exact matches only
Search in title
Search in content

Stay current with our monthly editions

Shopping cart
There are no products in the cart!
Continue shopping