Quantifying effects of irrigation and soil water content on electrical potentials in grapevines (Vitis vinifera) using multivariate statistical methods

Several studies have shown that physiological responses in plants, including fruit crops, are associated with changes in electrical potentials (EP), but it is often difficult to statistically quantify these responses. This study tested the effects of irrigation on EP in grapevines (Vitis vinifera), taking into account vapor pressure deficit (VPD) and position of electrodes along the stem by using multivariate analytical methods and a suite of statistical pretreatments. In two separate experiments, plants were exposed to one of two irrigation treatments in a greenhouse: (T1) irrigation once per day (Experiment 1), or no irrigation (Experiment 2); or (T2) irrigation three times or twice per day (Experiments 1 and 2, respectively). In each experiment, EP at three positions along the stem, soil (potting medium) water content, and VPD were continuously measured. In Experiment 2, stomatal conductance (gs) and stem water potential (SWP) were also measured for plants in each irrigation treatment as indicators of plant water status. Data were analyzed by Principal Component Analysis (PCA) to determine the effects of irrigation treatment on EP and difference in EP between pairs of electrodes (Delta EP) at various locations along the stem. Data were also analyzed by partial least squares (PLS) analysis to determine if EP or Delta EP could be used as predictors of changes in soil water content due to different irrigation treatments. Significant differences in soil water content due to irrigation treatments could be readily detected by difference in EP or Delta EP using PCA with Orthogonal Signal Correction pre-processing. Also, PLS showed that differences in soil moisture can be predicted by EP and/or Delta EP measurements at specific locations along the stem. Thus, the use of multivariate statistical methods was effective for relating EP and Delta EP measurements in grapevines to soil moisture due to differences in irrigation. (C) 2014 Elsevier B.V. All rights reserved.