Article Preview
TopIntroduction
The development of the Margaret River wine district (Western Australia) started in the mid-1960s under the impetus of agronomist John Gladstones who was among the first to identify the potential of the region for wine grape growing (Gladstones, 1965). The combination of a virtually frost free and high rainfall winter and a dry and warm summer was deemed ideal for high quality wine production. The Geographical Indication (GI), officially established in 1996, extends over 212,000 hectares for an area under vine of 5,800 hectares in 2016 (Margaret River Wine Association, 2018). The 80km-long Margaret River peninsula straddles two oceans: the Indian Ocean on its northern and western coastlines and the Southern Ocean along the southern coastline. This setting entails regional climate variations potentially impacting vine growing, and not necessarily captured by the current GI boundaries. In comparison, the Bordeaux wine district in France – also renowned for Cabernet sauvignon-based wines – covers 111,150 hectares for 39 recognised sub-regions, reflecting variations in vine growing conditions and wine styles (Conseil Interprofessionnel du Vin de Bordeaux, 2016; Institut national de l’origine et de la qualité, 2013).
According to Vaudour, Costantini, Jones and Mocali (2015) who studied the available literature on viticultural terroir zoning, the identification of variability in vine-growing areas has predominantly been investigated at a within-field scale as part of a precision viticulture approach (Bramley & Hamilton, 2007; Bramley, Ouzman, & Boss, 2011). The variability is usually identified by field measurements of oenological or viticultural parameters (grape composition, canopy characteristics, vine vigor or water status) and correlated to soil, topography or climate. Larger scale, regional to country-wide studies have focused on analyzing climate variability between wine regions within areas as large as Australia (Hall & Jones, 2010), New Zealand (Anderson, Jones, Tait, Hall, & Trought, 2012), Central Chile (Montes, Perez-Quezada, Peña-Neira, & Tonietto, 2012) or the western United States (Jones, Duff, Hall, & Myers, 2010).
The intermediate scale of district to regional variability has seen a recent interest with Geographic Information Systems (GIS) becoming more widespread and capable of processing larger amounts of geospatial data. Three areas have most notably been investigated: Europe (Fraga, Costa, & Santos, 2017; Herrera Nuñez, Ramazzotti, Stagnari, & Pisante, 2011; Moral, Rebollo, Paniagua, & García-Martín, 2016), the United States (Jones, Snead, & Nelson, 2004; Jones, Duff, & Myers, 2007; Nowlin & Bunch, 2016) and South Africa (Carey, 2005; Carey, Archer, Barbeau, & Saayman, 2009). For a majority of these publications, the methodology chosen involves the characterisation of the environmental aspects of a region through a GIS-based spatial analysis of climatic parameters or indices (often temperature or rainfall related) and topographic and soil variables. A limited number of studies have also included either field-based or remotely-sensed viticultural variables (Priori, Barbetti, Abate, Bucelli, & Storchi, 2014). Malone, Hughes, McBratney and Minasny (2014) focused on identifying terrons or soil-landscape units in the Lower Hunter Valley of New South Wales, Australia, based on soil and topographic considerations.