Wine Industry Newsletter

Cabernet genomic project – sensory findings

The Cabernet genomic project sought to determine the DNA sequence of 12 clones of Cabernet Sauvignon widely planted in Australia

Introduction

In 2014, the Western Australian wine industry secured a collaboration with Yalumba Nurseries and the Australian Wine Research Institute for a national project on clonal variation in wine grapes, led by the then Department of Agriculture and Food and the University of Western Australia (UWA).

Many important fruit crops, including grapevines, use vegetative propagation (cloning) to replant and expand plantings.

In the wine industry, the same flagship varieties have been commercially produced for centuries.  

This remarkable feature of the wine industry creates both problems and opportunities. While cloning perpetuates the variety and its marketable and/or agronomic qualities, somatic mutations arise spontaneously and accumulate with time through generations, resulting in phenotypic and agronomic differences and the creation of new clones, and ultimately new cultivars or trademarked clones.

Each mutation may be small – one single nucleotide – or large – a chromosome rearrangement or transposition of segments of a chromosome, including to other chromosomes.

Currently, there are no efficient means of selection, identification and assurance of novel clones.

The Cabernet genomic project sought to determine the DNA sequence of 12 clones of Cabernet Sauvignon widely planted in Australia, and identify DNA variants that can be used to develop markers, and lead research toward understanding the genetic basis of quality differences between clones.

The potential impact of this research is enormous, as currently many vineyards have little or no documentation of which clone or clones populate their vineyards. Being able to fingerprint and map the identity of a vineyard is just one opportunity. Being able to understand wine quality differences is another, which established the winemaking trials presented here, led by Research Officer Richard Fennessy.

Background

One of major populations of clones most interesting to the Western Australian industry is the ‘Houghton’ selection of Cabernet Sauvignon, initially established in the mid-late 1960s by the then Department of Agriculture and Food and Houghton Wines (Ian Cameron and Dorham Mann).

Individual vines were selected from the Swan Valley Houghton Vineyard (believed to have arrived in Australia from Cape Horn, South Africa) on the basis of ripening, yield and berry characters.

A selection of 20 or so were planted in foundation blocks in Gingin and the Great Southern for bulking up the vines.  These vines were the major source of the early plantings in Margaret River and Great Southern vineyards in the 1970s. Many are still in production.

Although it was quite common to plant ‘mass selections’ of the Houghton clones, i.e. a mixture of the top ~20, some vineyards established trials of individual clones. One of those was Gralyn Vineyard in Margaret River.

In the early 1970s, Gralyn was planted with one row each of Houghton #5, #13, #16 and SA126, which was sourced from South Australia and was the preferred clone from the late 1970s.

Since 2014, the Department of Agriculture and Food (now the Department of Primary Industries and Regional Development) have been collecting a range of field data from these rows and making small-lot batches of wine.

Unfortunately, records of which row was which clone were lost, however the genomics side of the project sought to establish their identity. Combined, this sensory analysis would provide a foundation for industry to re-capture the heritage of the Houghton clones.

Sensory analysis

Wines were made over three consecutive vintages and the winemaking protocol was consistent between years.

The wines were assessed on three occasions and conducted by a panel of 21, 28 and seven assessors, respectively in 2014, 2015 and 2016.

There were four sensory categories assessed: Aroma, Flavour, Tannins and Palate. Each category had different attributes assessed on an intensity scale in the range of 0-9.

The aroma‘s attributes included:  raspberry, cherry, cassis, currant, spice, vegetative and overall score.

Similarly, the flavour‘s attributes included: raspberry, cherry, cassis, currant, spice, plum, chocolate and overall score.

The palate had four attributes: weight, fruit sweetness, texture and length.

The tannins were assessed on presence/absence of the attributes soft, fine, powdery, drying, grippy and green. There was also an overall score out of 20.

The whole set of data, including all three years,  was analysed using Linear Mixed Models (LMMs) where each attribute for each category (Aroma, Flavour, Palate) was separately fitted as a response variable.

Overall the results from the analyses did not indicate many attributes with a significant difference between the clones. This does not necessarily indicate similarity of the clones, it may be attributed to other factors such as lack of training of the assessors or limitation due to design (lack of replication).

There were only two categories, aroma and flavour, where a significant difference between the clones for some of the attributes was observed as shown in Table1.  Raspberry was significant for both categories, while cassis was significant only for aroma, and currant and spice only for flavour.

Attribute vegetative, typical only for aroma, was also significant. Attribute currant in aroma was on the border of significance, P-value of 0.07.

Figures 1 and 2 illustrate the sum of the combined years of sensory data expressing intensity aroma and flavour attributes respectively.

Figure 1 Aroma intensities

Figure 2 Flavour intensities

Observations by clone

Row five

• Scored significantly higher in cassis aroma than rows six, seven and eight.
• scored significantly lower in raspberry aroma than row seven (highest) and significantly lower in  vegetative aroma and spice flavour than row six (highest).

Therefore this clone shows high cassis aroma with low vegetal notes and spice flavour.

Row six

• Scored significantly higher in currant aroma and flavour than row seven.
• Vegetal aroma and spice flavour scored the highest against the other clones.

This clone can be described as having dark fruit characters with spicy notes and varietal vegetal aroma.

Row seven

• Scored significantly higher than rows five and six for raspberry aroma and higher than row six for raspberry flavour.
• Scored significantly lower in cassis aroma than row five; currant aroma, vegetative aroma, currant flavour and spice flavour than row six.

Overall this clone appears to be dominated by red fruit characters.

Row eight

• Significantly lower in vegetative aroma and spice flavour than row six.

Overall this clone didn’t express dominance in any of the attributes, so unable to give a generalised description.

Conclusion

Now that these sensory data sets are compiled, the Department of Primary Industries and Regional Development and UWA are working hard to resolve the genetic identity, so the information can be integrated.

In addition to identity, the genomic research will begin to identify any DNA variation in genes known to determine some of the sensory characters described and build an understanding of ‘genes to wine’.