Identification and characterisation of terroirs depends on knowledge of environmental parameters, functioning of the grapevine and characteristics of the final product. A network of plots of Sauvignon blanc was delimited in commercial vineyards in proximity to weather stations at 20 localities and their viticultural and oenological response was monitored for a period of seven years. These experimental plots were further characterised with respect to climate, soil and topography. In order for this information to be of use within terroir zoning studies it had to be placed in a spatial context. This was achieved with the use of regression tree methodology, which determined the relative importance of the environmental and management related variables and regression trees for each dependent variable. A knowledge-driven model used the rules generated in the regression tree analyses to directly classify natural terroir units with respect to expected response of Sauvignon blanc in the Stellenbosch Wine of Origin District. The expected response of these terroir units was compared to data obtained from a separate network of Sauvignon blanc plots monitored during the 2005 harvest season.
Authors: Victoria A. CAREY (1), Eben ARCHER (1), Gérard BARBEAU (3) and Dawid SAAYMAN (3)
(1) Department of Viticulture and Oenology, Stellenbosch University, Private Bag X1, 7602 Matieland, South Africa
(2) Unité Vigne et Vin, Centre INRA d’Angers, 42 rue G. Morel, BP 57, 49071 Beaucouzé, France
(3) Distell, P.O. Box 184, 7599 Stellenbosch
Keywords: Sauvignon blanc, terroir, climate, soil, GIS