In the context of water management in agriculture, irrigation scheduling is critically important as it optimises water application to crops and can also target specific production goals. However, there is no consensus on the ideal irrigation scheduling strategy regarding crop water use efficiency (WUEc). In a premium Cabernet Sauvignon vineyard in Coonawarra, South Australia, over three growing seasons, irrigation scheduling strategies based on experience or historical knowledge (‘GROW’ treatment) were compared to data-driven strategies including crop evapotranspiration, and plant and soil water status thresholds to evaluate their effects on leaf- and vine-level WUEs. A final treatment, GROW + , that doubled the GROW level of irrigation was also evaluated in the third season. The WUE metrics were determined at the leaf, vine, and fruit scales as intrinsic WUE (WUEi), crop WUE (WUEc), and carbon isotope ratio (δ13C), respectively. Furthermore, the irrigation strategies were evaluated in the background of two contrasting soil types: Terra Rossa (light clay, well-drained) and Rendzina (heavier clay, poorly drained). Seasonal soil and vine water status, leaf gas exchange, and light interception were measured, and yield components and pruning weights were obtained following harvest. The amount of seasonal irrigation water based on the data-driven strategies was up to 65% lower across both soil types compared with the GROW or GROW + approaches. WUEi and δ13C were largely similar between treatments. However, for vines grown on Terra Rossa soil, little to no yield penalty was observed when data-driven irrigation scheduling was applied, in addition to increased WUEc values of up to 41%. It can be concluded that irrigation scheduling decisions based on data were superior to the conventional irrigation scheduling method on account of reducing irrigation water volume and increasing WUE, particularly in Terra Rossa soils.
Read full abstract