Comparing water potential variables under different water stress levels: a case study on Carménère grapevines
2019
Poblete-Echeverria, C. | Ahumada-Orellana, L. | Zuniga, M. | Sepulveda-Reyes, D. | Ortega-Farias, S.
Irrigation scheduling is a critical aspect of wine production, especially in some Mediterranean areas. One of the greatest difficulties of irrigation management is associated with poor or null estimations of plant water status. In this regard, water potential, measured by a pressure chamber is widely recognized as a practical tool to quantify plant water status under field conditions. However, diverse results have been presented in literature depending on the variable used to quantify the water potential [pre-dawn (Ψpd), leaf (Ψleaf) or stem (Ψstem)] the levels of water stress reached, as well as, the cultivar characteristics. The aim of this case study was to determine the suitability of the most used water potential variables (Ѱpd, Ѱstem, and Ѱleaf) for discriminating among plant water stress levels in grapevines 'Carménère'. To achieve this goal the vineyard was managed with four progressive regulated deficit irrigation (RDI) treatments during two consecutive growing seasons and the stomatal conductance (gs) was measured by an infrared gas analyzer (LI-6400) weekly. The entire data set was divided into three gs classes (mild, moderate, and severe water stress) according to the levels proposed in the literature for C3 plants. The suitability to discriminate among water stress levels was analyzed using a standard k-means clustering method evaluated with a confusion matrix approach. Ψleaf presented the highest variability within classes and the lowest overall accuracy in the classification (40%). Ψpd and Ψstem had a better classification performance, in particular, Ψstem technique presented a high sensitivity value for the class “Severe” (87%), which is an important aspect for an accurate water stress monitoring under drought conditions.
显示更多 [+] 显示较少 [-]