The specific objective of the study was to determine if the spatial variability of proximally sensed vineyard soil bulk EC would affect plant water status, and if this relation would affect leaf gas exchange, components of yield, berry composition, and flavonoids in both berries and wine.The study was conducted in a commercial vineyard in 2016 and 2017 with Cabernet Sauvignon grapevines grafted on 110R located in Healdsburg, CA, United States. In this vineyard, grapevines were planted at 1.83 m × 3.35 m . The grapevines were trained to a high quadrilateral, horizontally split trellis with two bilateral cordons. They were spur pruned with two buds per spur, and seven spurs per meter of the cordon. Irrigation was applied uniformly with a drip irrigation system, starting at fruitset to the end of veraison at 50% ETc . There were two emitters per grapevine, delivering 3.8 L·h −1 of water. Weather data was obtained from the California Irrigation Management Information System station to measure precipitation, air temperature, and reference evapotranspiration .An equidistant 33 m × 33 m grid with 35 experimental units was used for on-site measurements and berry samplings. Each experimental unit consisted of five plants. The locations of each central plant in these five plant experimental units were registered as the grid nodes with a GPS , large plastic garden pots wirelessly connected to a Trimble Pro 6T DGNSS receiver .Soil bulk EC was assessed with EM38 in 2016 when the vineyard soil was at field capacity condition. Both vertical dipole mode and horizontal dipole mode were used to assess EC at two depths, including deep soil and shallow soil .
The instrument was calibrated according to manufacturer instructions. The device was placed on a PVC sled and driven through the vineyard with an allterrain vehicle along the inter-rows. A distance of approximately 0.5 m from the vehicle to the device was maintained to avoid interference with the vehicle. A stratified grid was used to collect soil samples corresponding to the two depths at which we measured soil bulk EC. Soil texture was assessed according to the soil analysis method: hydrometer analysis in the North American Proficiency Testing program.Skin anthocyanins were analyzed by a reversed-phase HPLC consisting of a vacuum degasser, an autosampler, a quaternary pump, and a diode array detector with a column heater. A C18 reversed-phase column was utilized for analyzing anthocyanins. The mobile phase flow rate was 0.5 mL·min−1 , and two mobile phases were used, which included solvent A = 5.5% aqueous formic acid and solvent B = 5.5% formic acid in acetonitrile . The HPLC flow gradient started with 91.5% A with 8.5% B; 87% A with 13% B at 25 min; 82% A with 18% B at 35 min; 62% A with 38% B at 70 mins; 50% A with 50% B at 70.01 min; 30% A with 70% B at 75 min; 91.5% A with 8.5% B from75.01 min to 90 min. The column temperature was maintained at 25◦C. Detection of anthocyanins was carried out by the diode array detector at 520 nm. Wine proanthocyanidin subunits were characterized by acid catalysis in the presence of excess phloroglucinol by reversedphase HPLC . 1 mL of wine sample was applied to the Bond Elut C18 OH solid phase extraction cartridges to purify wine proanthocyanidins.
Eluents were evaporated and resuspended in 1 mL of methanol, and 0.25 mL methanolic extracts were combined with 0.25 mL of phloroglucinolysis reagent . The mixtures were then water bathed at 50◦C for 20 min. The reaction was stopped by mixing 200 µL of the sample mixtures with 1 mLof stopping reagent and then injected into the HPLC. The HPLC column consisted of two Chromolith RP-18e columns serially connected and protected by a guard column with the same material from EM Science . The mobile phase flow rate was 3.0 mL·min−1 . Two mobile phases were used, which included solvent A = 1% aqueous acetic acid and solvent B = 1% acetic acid in acetonitrile . The HPLC flow gradient started with 97% A with 3% B; 82% A, 18% B at 14 min; 20% A, 80% B at 14.01 min; 97% A, 3% B at 16.01 min until 20 min. All solvents used in this analysis were of HPLC grade, including acetonitrile, methanol, hydrochloric acid, and formic acid purchased from Fisher Scientific . Standards used for compound identification included malvidin 3-O-glucoside, -epicatechin purchased from Extrasynthese . Phloroglucinol was purchased from VWR .Geostatistical analysis was performed in the R language by using package “gstat” 1.1-6 . The bulk EC data were filtered by Tukey’s rule to remove outliers either below the first quartile by 1.5 inter-quartile range or above the third quartile by 1.5 inter-quartile range. To further remove the outliers, the data were filtered by the speed that the vehicle was driving, which was between 3.2 km per hour to 8.0 km per hour. Variograms were assessed by “automap” package 1.0-14 , and fitted to perform kriging. The soil bulk EC values were extracted from the location of each experimental unit, these values were further used to perform regression analysis. Kriging and k-means clustering on plant physiology variables were performed with the R packages “gstat” and “NbClust,” v3.0 . Universal kriging was utilized on plant water status because of the existing trend in longitude and latitude.
Variograms were assessed by “automap” package 1.0-14 , and fitted to perform universal kriging. The vineyard was delineated into two clusters by k-means clustering, including Zone 1 with higher water deficit and Zone 2 with lower water deficits. The separation described 78.1% in 2017 of the variability in the plant water status according to the result of between sum of squares/total sum of squares. The resulting maps were organized and displayed by using QGIS software . Cluster comparison was analyzed by “raster” package reported as Pearson’s Correlation between two cluster maps . Data were tested for normality by using Shapiro-Wilk’s test, and subjected to mean separation by using two-way ANOVA with the package “stats” in RStudio . Significant statistical differences were determined when p values acquired from ANOVA were <0.05, and the zones were classified according to Tukey’s honestly significant difference test. Regression analysis was performed by SigmaPlot 13.0 . Correlation coefficient between variables were calculated in by Pearson’s correlation analysis, and p-values were acquired to present the significances of the linear fittings.2◦Bx, 3.65 pH, 6.53 g·L −1 TA in 2017, and Zone 2 reached a TSS of 26.32◦Bx, 3.75 pH, 6.01 g·L −1 TA in 2016 and 23.71◦Bx, 3.58 pH, 7.22 g·L −1 TA in 2017. Before dividing the fruits from each zone into three dependent replicate fermentation vessels , the grapes were destemmed and crushed once transported into the winery. 50 mg·L −1 of SO2 was added to each vessel to prevent oxidation. Water was added to the musts to balance soluble solid level at 25◦Bx due to the highly possible stuck fermentation events may occur based on the high TSS levels. Dilution factors were considered when analyzing the final wine chemical composition. The must samples were inoculated with EC- 1118 yeast to initiate the fermentation in jacketed stainless steel tanks controlled by an integrated fermentation control system , and two volumes of must were pumped over twice per day by the system. The fermentations were carried at 25◦C until the residual sugar contents were below 3 g·L −1 . Malolactic fermentation was initiated with the addition of VinifloraR Oenococcus oeni at 12◦C and 60% humidity. The free SO2 levels were adjusted to 30 mg·L −1 after malolactic fermentation completed. Then the wines were sterile filtered and bottled before further chemical analysis. Wine samples were filtered by PTFE membrane filters and transferred directly into HPLC vials for anthocyanin analysis.Between the 2 years of the study, raspberry plant pot the precipitation amounts were different . The precipitation amount in the dormant season prior to 2016 was 559.5 mm . However, this amount was 898 mm in the 2016–2017 season. The precipitation during growing seasons in these 2 years were limited, there were only 51.6 mm of precipitation received in 2016 from April to harvest. In 2017, 107 mm of precipitation were received from April to harvest. The research site only received 11.1 mm in 2016 and 15.4 mm in 2017 during the study time in each year from June to harvest. There was a slight difference observed close to harvest . In 2016, GDD accumulation was 1183◦C at harvest . The GDD accumulation was greater in 2017 at 1220◦C by harvest . The cumulative ETo was greater in 2017 compared to 2016 . At harvest, the cumulative ETo was 750 mm in 2016, but it was relatively lower compared to 872.8 mm in 2017.In our previous work, we were unable to deduce a significant relationship between site topography variables such as absolute elevation and berry chemistry . Bramley et al. showed that soil bulk EC was directly related to soil clay content, which was contradictory to our findings. We attributed this discrepancy to the relatively stable soil texture throughout the season or even several seasons.
On the other hand, the effect of soil water content might be the major factor to influence plant development during the season. The soil texture and soil bulk EC sensing analysis conducted in this study were able to explain the variability in plant water status that the site topography could not. Soil texture and soil bulk EC can be related to spatial differences in soil water availability . Specifically, soil texture is a determinant of soil water holding capacity, hence affecting the amount of water available to the plants. In our study, the western section of the vineyard had greater loam proportion, where the grapevines were experiencing more severe water deficits . The eastern section had more sandy soil in both deep and shallow soil, where the grapevines were under less severe water deficits. Our findings are corroborated with previous work, where clay soil would lead to less plant available water, although clay soil had higher water holding capacity than sandy soil . Furthermore, Cabernet Sauvignon grapevines grown in clay soil would result in lower gs and An compared to grapevines grown in soils that had higher proportion of sandy soils . There was evident variability in soil bulk EC in this study. Previous studies reported that when soil bulk EC was proximally sensed, it was closely related to soil water content . We found that soil bulk EC was consistently and directly related to long-term 9 stem over the course of our study. Our findings are corroborated by previous works , where higher soil bulk EC values corresponded to higher soil water content. Previous studies suggested that the relationship between soil water content and soil bulk EC was soil-specific, and needed to include soil chemical and physical properties to explain variability and plant water status . Due to the limited amount of water put into wine grape vineyards, soil water content would be the major factor affecting soil electrical properties rather than the residual salinity after water evaporation from soil. The significant relationship between soil bulk EC and 9 stem in this study agreed with previous studies, indicating the possibility of soil bulk EC sensing being used to assess plant water status . Moreover, in our study, the spatial variability in grapevine physiology reflected the variability in soil bulk EC very well when assessed by proximal sensing. Due to the relationship of soil bulk EC on the amount of available water to plants reported in previous research , this approach had been utilized to identify the variability in the plant physiology based on the soil sensing technologies and apply targeted management strategies , and our study provided more evidence toward the feasibility of it.The variability we measured proximally in soil characteristics was reflected in plant water status and leaf gas exchange in our study. Previous research had reported that variable soil characteristics in space would cause spatial variations in plant water status . Although the precipitation amounts were vastly different between the two dormant seasons, the uniformly scheduled irrigation did not ameliorate the natural spatial variability in plant water status induced by soil properties. On the contrary, the separations in plant water status and leaf gas exchange were already significant even before the irrigation ceased after veraison. This proved that the spatial variability in the soil dominated the accessibility of the available soil water toward the plant, and made the spatial variability expressed in the grapevine.