Underpinning terroir with data: a quantitative analysis of biophysical variation in the Margaret River region of Western Australia

Abstract

Background and Aims The wine producers of Margaret River, as in many Australian wine regions, are interested in gaining a better understanding of the terroir of their region. Our aim was to determine whether the variation in soils and climate suggests that subdivision of the region may have merit. Methods and Results Clustering of digital map layers describing a range of climatic indices and soil properties of viticultural importance, using k‐means, enabled exploration of regional scale patterns of spatial variation. These methods are the same as those that have been successfully used at within‐vineyard scale in support of the development of Precision Viticulture. Whilst there is marked soil and climatic variation within the Margaret River geographical indication (GI), the results obtained from this analysis do not lend weight to previously proposed subdivisions of the Margaret River GI but rather, provide a guide for future targeted sensory analysis within the region. This was especially the case when the analysis was confined to just the 2.8% of the GI land area that is under vineyard. Conclusions Whilst the Margaret River GI shows marked biophysical variation, the data used in this analysis were not sufficiently strongly spatially structured at regional scale to indicate a basis for splitting the GI into subregions. However, some separation based solely on temperature and rainfall might form a useful basis for chemical and sensory analysis of wines and pursuit of understanding of the regional terroir. From both terroir and viticultural management perspectives, soil variation is likely of much greater importance at the within‐vineyard and whole of property scales, than at regional scale. Significance of the Study This is one of the first studies which uses purely data driven methods to examine biophysical variation within a wine region. Aside from being free of the bias that is introduced by heuristics, or examination of historically delineated terroirs, the work highlights the importance of focussing solely on the land which is under vine in evaluating terroir, rather than all of the land within a GI.

Underpinning terroir with data: on what grounds might subregionalisation of the Barossa Zone geographical indication be justified?

“…Similarly, the soil variation to the west of the hills which divide the Barossa and Eden Valleys (Figure 5) aligns closely with the seasonal variation (Figure 4c), but with the seasonal cluster identified in yellow in Figure 4c separated by the soil variation into the yellow and pale green clusters seen in Figure 6a. In other words, and in contrast to a similar study conducted in Margaret River (Bramley and Gardiner 2021) in which soil properties did not appear to contribute strongly to within-region terroir zoning, both AWC and CEC do appear to be important to potential terroir zoning within the Barossa. Of interest, neither inclusion of MJT (Figure 6b) nor of SRI (Figure 6c) contributed to the identification of patterns of within-region terroir variation compared to the analysis based solely on soil properties and the seasonal characterisation derived from GDD and GSR (Figure 6a).…”

“…Their results contrasted somewhat with a previous analysis based on a more traditional approach to land-use classification (Robinson and Sandercock 2014) and, more importantly, appeared to contrast with how local Barossa winemakers thought the GI could be split into so-called 'Grounds' whose wines might be perceived as distinctive. Somewhat similar work was recently conducted in the Margaret River GI of Western Australia (Bramley and Gardiner 2021) where only 2.8% of the GI land area is under vine. In the latter study, different results were obtained when viticulturally important soil and climate attributes were clustered just for land under vine, compared to when data for the entire GI were used as is common in terroir zoning research (e.g.…”

“…Accordingly, as in our earlier study (Bramley et al 2020), our focus was on the same climatic indices as used by Robinson and Sandercock (2014). With the exception of the addition of CEC, which was considered important as a surrogate indicator of soil fertility (Bramley et al 2020, Bramley and Gardiner 2021), in addition to AWC as an indicator of water availability, our consideration of soil was also similar. Nonetheless, this consideration of soil might be viewed as simplistic.…”

“…In contrast, Biss (2020) has recently concluded that topography has no bearing on wine quality in the Chablis region of France, when ‘quality’ is measured by the wine scores of judges and reviewers rather than by indices of wine composition. Similarly, Bramley and Gardiner (2021) considered topography as unlikely to be a contributor to a possible subregionalisation of the Margaret River region because there are no localised groupings of vineyards based on topography in that region and neither are its vineyards located at characteristic elevations. Collectively, these various results suggest that topography can be an important factor in considering terroir, with its importance varying on a case by case basis and across a range of scales – both horizontal and vertical.…”

“…Vaudour et al 1998, 2010, Jones et al 2004, Carey et al 2009, Bonfante et al 2011, 2018, Fraga et al 2017). These different results led to a different recommended basis for exploring within‐GI wine sensory and chemical differences than one offered on the basis of a ‘whole‐of‐region’ analysis (Lacorde 2019, Bramley and Gardiner 2021). Following our preliminary analysis of biophysical variation in the Barossa (Bramley et al 2020), it was of interest to explore this land use issue in a revised analysis given that only approximately 10% of the land in the Barossa GI is under vine.…”