Temperature variability assessment at vineyard scale: control of data accuracy and data processing protocol

Heatwaves or extreme heat events can be particularly harmful to agriculture. Grapevines grown in the Douro winemaking region are particularly exposed to this threat, due to the specificities of the already warm and dry climatic conditions. Furthermore, climate change simulations point to an increase in the frequency of occurrence of these extreme heat events, therefore posing a major challenge to winegrowers in the Mediterranean type climates. The current study focuses on the application of the STICS crop model to assess the potential impacts of heatwaves in grapevine yields over the Douro valley winemaking region. For this purpose, STICS was applied to grapevines using high-resolution weather, soil and terrain datasets over the Douro. To assess the impact of heatwaves, the weather dataset (1989-2005) was artificially modified, generating periods with anomalously high temperatures (+5 ºC), at certain onset dates and with specific durations (from 5 to 9 days). The model was run with this modified weather dataset and results were compared to the original unmodified runs. The results show that heatwaves can have a very strong impact on grapevine yields, strongly depending on the onset dates and duration of the heatwaves. The highest negative impacts may result in a decrease in the yield by up to -35% in some regions. Despite some uncertainties inherent to the current modelling assessment, the present study highlights the negative impacts of heatwaves on viticultural yields in the Douro region, which is critical information for stakeholders within the winemaking sector for planning suitable adaptation measures.

Measurement of carbon isotope discrimination in berry juice sugars at maturity (δ13C) provides an integrated assessment of water use efficiency (WUE) during the period of berry ripening, and when collected over multiple seasons can be used as an indication of drought stress response. Berry juice δ13C measurements were carried out on 48 different varieties planted in a common garden experiment in Bordeaux, France from 2014 through 2021 and were paired with midday and predawn leaf water potential measurements on the same vines in a subset of six varieties. The aim was to discriminate a large panel of varieties based on their stomatal behaviour and potentially identify hydraulic traits characterizing drought tolerance by comparing δ13C and hydroscapes (the visualisation of plant stomatal behaviour as a response to predawn water potential). Cluster analysis found that δ13C values are likely affected by the differing phenology of each variety, resulting in berry ripening of different varieties taking place under different stress conditions within the same year. We accounted for these phenological differences and found that cluster analysis based on specific δ13C metrics created a classification of varieties that corresponds well to our current empirical understanding of their relative drought tolerances. In addition, we analysed the water potential regulation of the subset of six varieties (using the hydroscape approach) and found that it was well correlated with some δ13C metrics. Surprisingly, a variety’s water potential regulation (specifically its minimum critical leaf water potential under water deficit) was strongly correlated to δ13C values under well-watered conditions, suggesting that base WUE may have a stronger impact on drought tolerance than WUE under water deficit. These results give strong insights on the innate WUE of a very large panel of varieties and suggest that studies of drought tolerance should include traits expressed under non-limiting conditions.

The main innovation in the VINIoT service is the joint use of two technologies that are currently used separately: vineyard monitoring using multispectral imaging and deployed terrain sensors. One part of the system is based on the development of artificial intelligence algorithms that are feed on the images of the multispectral camera and IoT sensors, high-level information on water stress, grape ripening status and the presence of diseases. In order to obtain algorithms to determine the state of ripening of the grapes and avoid losing information due to the diversity of the grape berries, it was decided to work along the first year 2020 at berry scale in the laboratory, during the second year at the cluster scale and on the last year at plot scale. Different varieties of white and red grapes were used; in the case of Galicia we worked with the white grape variety Treixadura and the red variety Mencía. During the 2020 and 2021 campaigns, multispectral images were taken in the visible and infrared range of: 1) sets of 100 grapes classifying them by means of densimetric baths, 2) individual bunches. The images taken with the laboratory analysis of the ripening stage were correlated. Technological maturity, pH, probable degree, malic acid content, tartaric acid content and parameters for assessing phenolic maturity, IPT, anthocyanin content were determined. It has been calculated for each single image the mean value of each spectral band (only taking into account the pixels of interest) and a correlation study of these values with laboratory data has been carried out. These studies are still provisional and it will be necessary to continue with them, jointly with the training of the machine learning algorithms. Processed data will allow to determine the sensitivity of the multispectral images and select bands of interest in maturation.

Making high quality but affordable Pinot noir (PN) wine is challenging in most terroirs and New Zealand’s (NZ) situation is no exception. To increase the probability of making highly typical PN wines producers choose to grow grapes in cool climates on lower fertility soils while adopting labour intensive practices. Stringent yield targets and higher input costs necessarily mean that PN wine cost is high, and profitability lower, in line-priced varietal wine ranges. To understand the reasons why higher yielding vines are perceived to produce wines of lower quality we have undertaken an extensive study of PN in NZ. Since 2018, we established a network of twelve trial sites in three NZ regions to find individual vines that produced acceptable commercial yields (above 2.5kg per vine) and wines of composition comparable to “Icon” labels. Approximately 20% of 660 grape lots (N = 135) were selected from within a narrow juice Total Soluble Solids (TSS) range and made into single vine wines under controlled conditions. Principal Component Analysis of the vine, berry, juice and wine parameters from three vintages found grape berry mass to be most effective clustering variable. As berry mass category decreased there was a systematic increase in the probability of higher berry red colour and total phenolics with a parallel increase in wine phenolics, changed aroma fraction and decreased juice amino acids. The influence of berry size on wine composition would appear stronger than the individual effects of vintage, region, vineyard or vine yield. Our observations support the hypothesis that it is possible to produce PN wines that fall within an “Icon” benchmark composition range at yields above 2.5kg per vine provided that the Leaf Area:Fruit Weight ratio is above 12cm2 per g, mean berry mass is below 1.2g and juice TSS is above 22°Brix.

A large varietal collection including over 1700 varieties was maintained in Conegliano, ITA, since the 1950s. Phenological data on a subset of 400 grape varieties including wine grapes, table grapes, and raisins were acquired at bud break, flowering, veraison, and ripening since 1964. Despite the efforts in maintaining and acquiring data over such an extensive collection, the data set has varying degrees of missing cases depending on the variety and the year. This is ubiquitous in phenology datasets with significant size and length. In this work, we evaluated four state-of-the-art methods to estimate missing values in this phenological series: k-Nearest Neighbour (kNN), Multivariate Imputation by Chained Equations (mice), MissForest, and Bidirectional Recurrent Imputation for Time Series (BRITS). For each phenological stage, we evaluated the performance of the methods in two ways. 1) On the full dataset, we randomly hold-out 10% of the true values for use as a test set and repeated the process 1000 times (Monte Carlo cross-validation). 2) On a reduced and almost complete subset of varieties, we varied the percentage of missing values from 10% to 70% by random deletion. In all cases, we evaluated the performance on the original values using normalized root mean squared error. For the full dataset we also obtained performance statistics by variety and by year. MissForest provided average errors of 17% (3 days) at budbreak, 14% (4 days) at flowering, 14.5% (7 days) at veraison, and 17% (3 days) at maturity. We completed the imputations of the Conegliano dataset, one of the world’s most extensive and varied phenological time series and a steppingstone for future climate change studies in grapes. The dataset is now ready for further analysis, and a rigorous evaluation of imputation errors is included.