Assessing clonal variability in Chardonnay and Shiraz for future climate change
Abstract
The aim of this project was to assess the viticultural performance and wine sensory properties of a common selection of Chardonnay and Shiraz clones growing in a range of diverse climates in Australia. Across vintages there were statistically significant differences in small-lot wines between clones within a region, and between regions. The differences in wine descriptors between regions give an insight to how wine styles within each of these regions may change in a warmer climate.
However, there were no consistent trends in sensory differences between clones. At regional tastings participants demonstrated clear preferences for individual clones, but these too varied across regions and between seasons. These outcomes indicate that there is merit in planting a range of clones within vineyards and then using the diversity between the wines for winemakers to determine the specific end-product for the marketplace. Preliminary Shiraz genomics work detected dissimilarities between the clones DNA, and the Australian clones formed a group distinct from the three Shiraz clones recently imported from France.
Summary
Chardonnay and Shiraz represent the two major wine grape varieties in Australia. Both varieties are produced across a wide range of climates for different wine styles and price points. Over the last decade, an increasing number of clones of each variety have been planted, sometimes with limited knowledge of their performance and impact on wine attributes. In Australia, historical clonal evaluation has used yield as the main selection criteria, underpinned by industry goals to improve productivity and reduce seasonal yield variation. This project was developed to evaluate grapevine clones through to winemaking and comprehensive sensory assessment, with a view to making resilient and superior planting material available to the industry.
Mature plantings of Shiraz and Chardonnay in Victoria, South Australia and Western Australia were selected that comprised a range of clones, some of which were common to all sites. In choosing these sites, we also sought diverse climates to test our hypothesis that differences in climate between sites could be used as a surrogate for climate change. Our goal was to explore if the sensory attributes of a currently cool region could gradually become similar to those of a currently warm region and, similarly, will the profile of wines from a currently warm region become more like those of a currently hot region into the future? By assessing the fruit from common clones across multiple regions, the interaction between climatic conditions and clone could be investigated.
Suitable sites of Shiraz and Chardonnay were identified in the Riverland and Barossa in South Australia; the Grampians and Drumborg in Victoria, and Margaret River and the Great Southern in Western Australia. Field work commenced in the 2013-14 season. Three consecutive seasons of vine performance and wine sensory data were collected from the four Shiraz sites, and four consecutive years of data from the five Chardonnay sites. The majority of the sites were not replicated clonal trials, rather being row-by row plantings of the different clones, which limited the statistical analysis of the data sets. All of the clones, with the exception of perhaps the Gingin Chardonnay clone, had previously been evaluated for yield performance, so this was not the primary focus.
Heat summation and phenology were recorded at each site and used to predict the future harvest date in a warmer climate. The impact of a 0.5, 1.0. 1.5 or 2.0oC increase in daily temperature between budburst and harvest was modelled using current daily heat summation required to ripen both Chardonnay and Shiraz to the target maturity required for small lot winemaking. The modelling indicated an earlier harvest date for both varieties at all sites as well as vintage compression.
All wines were made using identical protocols and the same panel conducted the sensory descriptive analysis each year. Statistical analysis of sensory scores revealed significant differences between clonal wines for most of the attributes (approximately 30) in each season. Principal Component Analysis (PCA) was performed using combined data from all clones at the regional level and for all clones within each region, for all vintages. PCA plots at the regional level often resulted in distinct separation between the regions. For example, the 2017 Chardonnay wines from each region occupied a separate quadrat of the PCA plot, with each quadrat containing a completely different set of descriptors characterising the wines from the cool to warm regions. In other cases, wines tended to group together. For example, Chardonnay wines from Margaret River and the great Southern were grouped in the same quadrat for each of the four vintages, suggesting that they were more similar to each other than the other three regions, whilst Riverland Chardonnay wines were in a separate quadrat for three of the four years. In 2016 the four Shiraz regions separated into each of the four quadrats of the PCA plot and in the other two years, wines from the Barossa were separated from all other regions.
Shiraz sensory descriptors were used in each of the three years to determine if there were any common descriptors for each region. Barossa wines were characterised as having dark fruit and sweet spice aroma and high opacity, with Grampians wines having consistent red fruit aroma, floral and confection. Margaret River wines had consistent red fruit aroma. Riverland wines showed considerable variation across the three years, with no consistent traits. PCA plot comparison between vintages was not possible due to the reduced number of sensory attributes that were common across years (especially for the Chardonnay wines).
Overall, the PCA plots give some insight into how regional wine styles may change in a warming climate. At the regional level, the PCA plots separated the clones either within a quadrat or in some cases into different quadrats of the plot, indicating that there were differences in the sensory profile of the clones. In some cases these differences may have been due to maturity at harvest (using wine alcohol concentration as an index). However, there were a number of examples where clonal wines of similar alcohol concentration had distinct sensory profiles. While the regional PCA plots revealed differences between clonal wines in each season, there did not appear to be any consistent pattern in this separation between seasons.
Statistical analysis of the pooled sensory data within a region was used to compare regions and similarly within each region to compare clones. Radar plots, each with Least Significant Difference values included, were constructed using these data sets. There were significant differences in the majority of the sensory scores between regions in each year, however, the much smaller set of sensory attributes that were common to all years did not permit reliable statistical analysis of the effect of season on scores. Within each region there were again significant differences in sensory scores between clones in each year, but the number of attributes that were significantly different varied between regions. For example, for the 2015 Margaret River Chardonnay wines there were 14 significantly different sensory traits compared with seven for the 2017 wines.
Project results were presented to industry through many regional wine tastings, both in the project regions and elsewhere, and these were always well-attended and well-received. At all workshops we asked participants to rank their preferences, based on their overall perception of the wines and their potential as finished commercial wines. At most workshops, there was a most preferred clonal wine, and rankings of the same selection of wines tended to differ between regional groups. Although the method was different from the descriptive analytical approach used by the AWRI sensory panel, both approaches demonstrate that there were clonal differences between the wines.
The geographically diverse origins of the Shiraz clones used in this work presented the opportunity for a limited investigation of the level of genetic difference between the clones. Phylogenetic relationships between the seven Australian Shiraz clones and three French clones were determined. The phylogenetic tree arranged the French and the Australian-based clones into two distinct and separate groups (clades). The shared nucleotide variants of the Australian clones suggested a common origin for these clones which predates establishment of clonal selection programs and potentially also their original importation to Australia. Despite their shared heritage, the Australian clones appear quite distinct from each other, although they exhibited less intra-clade diversity than the more recently-imported French Shiraz clones.
The origin, selection history and published data on yield performance of all Chardonnay and Shiraz clones used in this study was researched and documented. The clones were also tested for the range of grapevine viruses for which diagnostic protocols are available in Australia. Overall, the clones were remarkably free of virus and virus-like diseases.
Wines of both varieties were presented at multiple workshops across four States for local winemaker/viticulturist assessment, as described above.
Preliminary results were presented in a plenary session and workshop at the 2016 Australian Wine Industry Technical Conference and a workshop at the 2019 Conference. Two presentations were made to the Californian Industry in late 2018.