Assessment of the nitrogen status of grapevines with high spatial, temporal resolution offers benefits in fertilizer use efficiency, crop yield and quality, and vineyard uniformity. The primary objective of this study was to develop a robust predictive model for grapevine nitrogen estimation at bloom stage using high-resolution multispectral images captured by an unmanned aerial vehicle (UAV). Aerial imagery and leaf tissue sampling were conducted from 150 grapevines subjected to five rates of nitrogen applications. Subsequent to appropriate pre-processing steps, pixels representing the canopy were segmented from the background per each vine. First, we defined a binary classification problem using pixels of three vines with the minimum (low-N class) and two vines with the maximum (high-N class) nitrogen concentration. Following optimized hyperparameters configuration, we trained five machine learning classifiers, including support vector machine (SVM), random forest, XGBoost, quadratic discriminant analysis (QDA), and deep neural network (DNN) with fully-connected layers. Among the classifiers, SVM offered the highest F1-score (82.24%) on the test dataset at the cost of a very long training time compared to the other classifiers. Alternatively, QDA and XGBoost required the minimum training time with promising F1-score of 80.85% and 80.27%, respectively. Second, we transformed the classification into a regression problem by averaging the posterior probability of high-N class for all pixels within each of 150 vines. XGBoost exhibited a slightly larger coefficient of determination (R² = 0.56) and lower root mean square error (RMSE) (0.23%) compared to other learning methods in the prediction of nitrogen concentration of all vines. The proposed approach provides values in (i) leveraging high-resolution imagery, (ii) investigating spatial distribution of nitrogen across a vine’s canopy, and (iii) defining spatial zones for nitrogen application and smart sampling.

Grapevine productivity, and berry and wine flavonoid concentration, depend on the interactions of cultivar, environment, and applied cultural practices. We characterized the effects that mechanical leaf removal and irrigation treatments had on the flavonoid concentration of ‘Merlot’ (Vitis vinifera, L.) grape berries and wines in a hot climate over two growing seasons with contrasting precipitation patterns. Leaves were removed by machine, either at prebloom (PBLR), or at post-fruit-set (PFLR), or not removed (control) and irrigation was either applied as sustained deficit irrigation (SDI) at 0.8 of crop evapotranspiration (ET_c) from budbreak to fruit set, or regulated deficit irrigation (RDI) at 0.8 ET_c from bud break to fruit set, 0.5 ET_c from fruit set to veraison, and 0.8 ET_c from veraison to harvest, of ET_c In 2014, PFLR reduced the leaf area index (LAI) compared to control. The RDI decreased season-long leaf water potential (Ψ_Int) compared to SDI. However, in 2015, none of the treatments affected LAI or Ψ_Int. In 2014, berry flavonoid concentrations were reduced by PBLR as well as SDI. SDI increased the flavonoid concentrations in wine, and PFLR increased some wine flavonols in one season. No factor affected the concentrations of wine proanthocyanidins or mean degree of polymerization. Thus, mechanical PFLR and RDI may increase berry flavonoid accumulation without yield reduction, in red wine grapes cultivars grown in hot climates when precipitation after bud break is lacking. However, spring precipitation may influence the effectiveness of these practices as evidenced by this work in a changing climate.