Understanding Spatiotemporal Patterns of Chemical Attributes in ‘Vitis vinifera’ L. cv. Cabernet Sauvignon Vineyards in Central California as a Basis for Predicting Fruit Composition

Abstract

Spatial variability of vine productivity in winegrapes is important to characterise as both yield and quality are relevant for the production of different wine styles and products. Few studies have analysed spatial variability of individual fruit compositional attributes, and even fewer in Vitis vinifera L. cv. Cabernet Sauvignon in California, USA. Previous studies have focused on basic chemistry (pH, TA, TSS), groups of attributes (total phenolics), or fruit colour, and few have reported maps of spatial variability of individual aroma precursors or specific phenolic compounds related to mouthfeel in the resulting wines. The overall objectives of the research presented in this thesis were to understand how patterns of variability of Cabernet Sauvignon fruit composition changed over time and space, how these patterns could be characterised with proximal and remote measurements, and how spatial patterns of the variation in specific fruit compositional attributes can aid in improving management decisions. Prior to the 2017 vintage, 125 data vines were distributed across each of four vineyards in the Lodi American Viticultural Area (AVA) of central California. Each data vine was sampled at commercial harvest in 2017, 2018, and 2019. Yield components and fruit composition were measured at harvest for each data vine, and maps of yield and fruit composition were produced for eight ‘objective measures of fruit quality’: total anthocyanins, polymeric tannins, quercetin glycosides, malic acid, yeast assimilable nitrogen, β-damascenone, C6 alcohols and aldehydes, and 3-isobutyl-2-methoxypyrazine. Maps were produced for each compound in each vineyard to assess the temporal stability of their patterns of spatial variability, and to identify which compounds were most useful in describing overall fruit compositional variability. Of all the compounds analysed, patterns of variation in anthocyanins and phenolic compounds were found to be most stable over time. Given this relative stability, management decisions focussed on fruit quality could be based on zonal descriptions of anthocyanins or phenolics to increase profitability in some vineyards. In addition to the yield and fruit composition measurements in each season, dormant season pruning weights and soil cores were collected at each location. In each vineyard, elevation and soil apparent electrical conductivity surveys were completed, and remotely sensed imagery was captured by fixed wing aircraft and two satellite platforms at major phenological stages. The data collected were used to develop relationships among biophysical data, soil, imagery, and fruit composition. Remote sensing measures provided similar patterns of variability to those obtained by ground measures. Characterisation of patterns of spatial variability is difficult because of the cost associated with large sampling numbers and densities required to produce geostatistically rigorous maps. The standardised and aggregated samples from four vineyards over three seasons were included in the estimation of ‘common variograms’ to assess how this technique could aid growers in producing geostatistically rigorous maps of fruit composition variability without cumbersome, single season sampling efforts. Overall, the characterisation of spatial variability of multiple fruit composition parameters is important for the development of prescriptive farming practices aimed at the enhancement of wine quality.