Abstract
A field study was conducted in central California to characterize the effects of rootstock genotype and amounts of applied water on the productivity and anthocyanin composition of Zinfandel (Vitis vinifera L.) grape berry in a hot climate. Zinfandel grafted onto either Freedom (Fresno 1613–59 × Dogridge 5, 27% V. vinifera hybrid; high vigor, nematode resistant) or Salt Creek (V. champinii; high vigor, phylloxera and nematode resistant, and salt tolerant) rootstock was studied during two growing seasons under sustained deficit irrigation (SDI) and regulated deficit irrigation (RDI). Midday leaf water potential (Yl), canopy architecture, yield, berry composition, and berry skin anthocyanin concentration were measured at harvest. The Yl of Zinfandel was consistently higher with SDI, while rootstock genotype did not affect it. Zinfandel grafted on Freedom consistently had greater berry weight, cluster number and weight, and yield per vine compared to Salt Creek. The seed number per berry and skin and seed mass of Zinfandel were greater when grafted on Freedom compared to Salt Creek. There were few effects of irrigation regimen or rootstock on berry composition in either year. Total berry skin anthocyanin concentration of Zinfandel was consistently greater with SDI than RDI and greater with Freedom in the second year. The tri-hydroxylated anthocyanin proportion of Zinfandel was consistently greater with Freedom rootstock. RDI reduced the water footprint of Zinfandel regardless of rootstock, but the associated decrease in yield was commercially unacceptable. Our results suggest that SDI in combination with Freedom rootstock can enhance water productivity based on limited reductions in yield combined with higher anthocyanin concentrations in berry skin in a resource-limited area.
Most vineyards in the San Joaquin Valley (SJV) of California experience water stress due to seasonal droughts, and the number of warm years and lengths of seasonal drought are increasing (Cook et al. 2015). Water deficits can limit grapevine yield and alter berry composition (Chaves et al. 2010). The grapevine has various physiological and morphological adaptations enabling it to produce a crop under conditions of water deficit (Koundouras et al. 2009).
Research emphasis has recently focused on water productivity, which is the amount of marketable product produced per unit of water consumed in evapotranspiration (ETc; Fereres and Soriano 2007, Chaves et al. 2010). Deficit irrigation of winegrapes increased water productivity over full irrigation. However, the increased water productivity was accompanied by reduced yield (Chaves et al. 2010). Vine response to applied water amounts depends on the pattern and severity of the imposed water stress (Hardie and Considine 1976). This knowledge was used to develop deficit irrigation strategies. In one strategy, sustained deficit irrigation (SDI), a uniform fraction of ETc is applied throughout the growing season and vine water stress progressively increases through depletion of soil water reserves (Cook et al. 2015). In another strategy, regulated deficit irrigation (RDI), a water stress is imposed on the vine at particular phenological stages, usually pre- or postveraison, and then altered during another phenological stage (Kennedy et al. 2002, Terry and Kurtural 2011). Water deficits were shown to consistently promote higher concentrations of anthocyanins and flavonols in red wine grapes (Kennedy et al. 2002, Terry and Kurtural 2011). Growth of berries was inhibited more and concentrations of flavonoids in berry and wine increased when water deficits were imposed before rather than after veraison (Matthews and Anderson 1989). Only postveraison water deficits inhibited fruit growth in Cabernet Sauvignon (Kennedy et al. 2002) and Syrah (Terry and Kurtural 2011). Gene expression studies investigating the regulation of flavonoid biosynthesis in grapevine concluded that both pre- and postveraison water deficits directly increased gene expression and accumulation of anthocyanins. In Merlot, anthocyanin concentration increased primarily due to a direct response to water deficit leading to overexpression of flavonoid synthesis genes, in particular UDP-glucose:flavonoid 3-O-glucosyltransferase (UFGT), chalcone synthase (CHS2, CHS3), glutathione S-transferase (GST), and flavonoid 3′-hydroxylase (F3H) (Castellarin et al. 2007).
Most winegrape vineyards in the SJV have been grafted on rootstocks. Rootstocks are used to provide resistance or tolerance to phylloxera and nematodes, adverse soil conditions such as high or low pH, drought, and salt (Berdeja et al. 2014). Rootstocks also affect scion phenology, leaf area, canopy development (Koundouras et al. 2009), pruning weight, and yield (Stevens et al. 2008). Under nonirrigated conditions, rootstocks with different water stress sensitivity may affect berry anthocyanin composition through their effect on carbohydrate availability (Ezzahouani and Williams 2007). Rootstock genotype did not affect berry growth and the concentration of berry skin phenolics (Koundouras et al. 2009).
Freedom (Fresno 1613–59 × Dogridge 5, 27% V. vinifera hybrid) is a widely used rootstock, particularly in SJV vineyards where root-knot nematodes are the key soilborne pest (Garris et al. 2009). In the 45 years since its release, Freedom and its half-sibling Harmony have largely replaced 1613C and Dogridge in new plantings because of their nematode resistance and improved horticultural characteristics. Salt Creek (V. champinii) is also used in the SJV, where root-knot nematode and phylloxera tolerance are needed (Christensen 2003). Salt Creek performed better in sandy, low-vigor soils and has good salt tolerance (Christensen 2003). However, there is lack of knowledge on scion performance in response to these widely planted rootstocks in the SJV.
In the present study, we analyzed the impact of two water stress treatments (SDI and RDI) on two rootstocks (Freedom and Salt Creek), both grafted with Zinfandel clone 1A. The objective of this study was to determine the effect of rootstocks and water stress treatments on productivity, berry composition, and berry skin anthocyanin concentration of Zinfandel in a hot climate growing region.
Materials and Methods
Site description and experimental design
This study was conducted at a commercial vineyard in Kern County, CA (lat. 35.03°N; long. 118.92°W; elevation 137 m, zero to two percent slope). The vineyard was planted with Zinfandel clone 1A on two rootstocks described below at 2.3 m × 3.35 m (vine × row) spacing in a north to south orientation. The experimental vineyard was planted in 2008 and treatments described below were applied in two consecutive years with field-grown plants in 2013 and 2014. The grapevines were trained on a bilateral cordon at 1.4 m above ground level with two support wires at 1.70 m and a 20 cm T-top. The soil type was Premier sandy-loam soil, a coarse-loamy, mixed, superactive, calcareous thermic Xeric Torriorthent derived from granitoid parent material (www.nrcs.usda.gov/).
The vineyard had been mechanically pruned to a 100 mm spur height with a 600 mm-wide, California-sprawl pruner to a node density of 55 nodes per meter of row since 2011. The vineyard was drip-irrigated with pressure-compensating emitters delivering 1.89 L/hr. Irrigation had been managed as follows since 2011. Vineyard crop evapotranspiration (ETc) was estimated as the product of reference evapotranspiration (ETo) and seasonal crop coefficients (Kc) (Allen et al. 1998). The ETo was obtained from the California Irrigation Management Information System (CIMIS; www.cimis.water.ca.gov/) weather station (#125) in Arvin, CA. The amount of precipitation received and the additional irrigation amounts were recorded weekly. The seasonal Kcs used to schedule irrigation at this site were developed by measuring shade cast on the vineyard floor beneath the canopy of vines irrigated at 0.8 ETc (SDI) treatments at solar noon weekly. The shaded area beneath the canopy was determined by counting the number of equidistant 0.01 m2 cells on an 18 m2 grid and summing their area.
The growing degree days (GDD) were calculated with a threshold of 10°C with data obtained from CIMIS. All other cultural practices were carried out according to commercial industry standards for that area (Bettiga 2013). The experiment was a two (irrigation treatment) × two (rootstock) factorial arrangement of treatments in a randomized complete block with four replications. Each experimental unit consisted of 357 vines, of which 32 were measured.
Rootstock treatments
Rootstocks assessed in this experiment included Freedom (Fresno 1613–59 × Dogridge 5, 27% V. vinifera hybrid) and Salt Creek (V. champinii).
Irrigation treatments
There were two irrigation treatments. A control treatment of SDI at 0.8 estimated ETc was applied from anthesis until harvest at modified Eichhorn-Lorenz (E-L) stage 38 with a midday leaf water potential (Ψl) threshold of −1.2 MPa. An RDI treatment was applied at 0.8 ETc from anthesis to fruit set (E-L stage 28) with a Ψl threshold of −1.2 MPa, at 0.5 ETc from fruit set to veraison (E-L stage 35) with a Ψl threshold of −1.4 MPa, and at 0.8 ETc from veraison until harvest with a Ψl threshold of −1.2 MPa. Irrigation treatments in each year were not initiated until Ψl reached −1.0 MPa for vines in the 0.8 ETc treatments. The water status of the grapevines throughout the growing season was monitored weekly starting at the 80th day of the year by measuring the Ψl. For each treatment-replicate, four fully expanded leaves exposed to the sun, showing no sign of disease or damage, were selected. A zip-top plastic bag was placed over a single leaf and sealed before the petiole was excised to suppress transpiration. Ψl was then directly determined using a pressure chamber (Model 610 Pressure Chamber Instrument; PMS Instrument Co., Corvallis, OR).
Canopy architecture and microclimate
Canopy architecture and microclimate variables were assessed by measuring exterior and interior leaf contact numbers and percentage canopy gaps in the fruiting zone with 10 insertions per data vine at 15 cm intervals at 150 cm above ground level at fruit set in each year of the study as described elsewhere (Wessner and Kurtural 2013). A ceptometer (AccuPAR-80; Decagon Devices, Pullman, WA) was placed directly above the cordon, within the fruiting zone and parallel to the vine row at the head of each grapevine, oriented due south. Four measurements were taken with the ceptometer from 32 vines per experimental unit. Ambient readings were taken 60 cm above the canopy surface. The remaining three measurements were taken within the fruiting zone at the head of the vine. Measurements were taken at midday with photosynthetically active radiation (PAR) values ranging ~1700 to 1900 μmol/m2/s. The three fruiting zone PAR measurements were combined and expressed as the percentage total ambient PAR for the daylight period. Leaf area was measured destructively at modified E-L stage 35 by defoliating a one-meter section of canopy. The number of shoots per meter was measured. Leaf area of a one-meter section of canopy was measured with the Li-Cor 3100 Leaf Area Meter (Li-Cor Environmental, Lincoln, NE) and extrapolated to calculate leaf area per vine as described (Cook et al. 2015).
Yield components, efficiency, and water footprint
Yield data was collected in 2013 and 2014. Harvest commenced when berry total soluble solids reached 20 Brix because the vineyard was under contract to harvest at that maturity. Each treatment replicate was harvested manually and the clusters harvested per vine were counted and weighed with a CGI-70 top-loading scale (CGI Scale Co., Ventura, CA). Cluster mass was calculated by dividing weight of clusters harvested by the number of clusters. One hundred berries were randomly collected and weighed on an analytical top-loading digital scale (Mettler Toledo ML-104, Columbus, OH). Leaf area to fruit ratio was calculated by dividing the leaf area/vine by the yield/vine (m2/kg). The water footprint (m3/ton) was calculated as the ratio of volume water applied per unit crop harvested. The amount of water applied was calculated by summing the volume of precipitation received and irrigation water applied from 15 March to 1 Nov in each year.
Berry composition
The berry total soluble solids percentage (Brix), juice pH, and titratable acidity (TA, as g tartaric acid/L) were determined from a random 100-berry sample collected at harvest from each treatment replicate. The Brix was measured using a digital refractometer (PR-32 Palette digital refractometer; ATAGO USA, Bellevue, WA). The juice pH was determined with a glass electrode and an Accumet 13-620-183A AB15 pH meter (Fisher Scientific, Pittsburgh, PA). TA was measured by titrating to an endpoint pH of 8.2 with 0.1 N sodium hydroxide and expressed as g/L.
Berry skin anthocyanin composition
The anthocyanin concentration of berry skin was determined using an exhaustive extraction method (Cook et al. 2015). At harvest, 20 random berry samples were collected from each treatment replicate, weighed, frozen in liquid nitrogen, and stored at −80°C until analyzed. Berry skins were manually removed, rinsed with deionized water, and counted. Samples were lyophilized (Triad Freeze Dry System; Labconco, Kansas City, MO) and reweighed to obtain the dry weight. The average dry weight of single berry skin per treatment-replicate was determined by dividing the total lyophilized skin mass by the number of berry skins. The samples were extracted in 20 mL 66% (v/v) acetone solution in darkness for 24 hr. Samples were filtered under vacuum, the grape marc was discarded, and a 1-mL sample was collected. The acetone was evaporated from the samples under vacuum with a centrivap (model 7810010, Labconco) attached to a −103°C cold trap (model 7385020, Labconco) and brought up to a volume of 5 mL. Samples were centrifuged for 15 min at 1400 g. The supernatant was then pipetted into a 2-mL HPLC vial and subjected to HPLC-DAD analysis.
HPLC analysis
Anthocyanin analysis was conducted using reversed-phase high-performance liquid chromatography (HPLC) using an Agilent 1100 series (Santa Clara, CA) modular HPLC system. The HPLC system included a system controller, G1379A degasser, G1311A quaternary pump, G1313A autosampler, G1316A column compartment, and a G1315A DAD/UV-vis detector. Data were processed using ChemStation v. B.04 designed for an LC system. Separation of anthocyanin compounds was performed with a LiChrospher 100 RP-18 (4 × 250 mm, 5 μm particle size) column (Agilent); a guard column of the same material was also installed and the column temperature was maintained at 40°C.
Three mobile-phase solutions were used for analysis. The solvents were (A) 50 mM ammonium dihydrogen ammonium phosphate adjusted to pH 2.6, (B) 20% Mobile A + 80% acetonitrile (v/v), and (C) 0.2 M orthophosphoric acid adjusted to pH 1.5. Solvents established the gradient as reported elsewhere (Cook et al. 2015). Analytical grade water was purified in-house with a Labostar Ultrapure Water System (Siemens, Warrandale, PA) before use. The mobile phase components were of HPLC grade and were purchased from Fisher Scientific. Spectra were recorded from 190 to 600 nm.
Quantification of anthocyanins was conducted by measuring peak area at 520 nm. The commercial standard used was malvidin-3-O-glucoside (Extrasynthése, Genay, France). Individual anthocyanin compounds were tentatively identified according to order of elution, retention times of pure compound, and previous research (Cook et al. 2015). Total skin anthocyanins (TSA) was calculated by summing identified individual anthocyanin compounds.
Statistical analysis
Data for all parameters were tested to verify whether the assumptions of analysis of variance (ANOVA) were met using Shapiro-Wilk’s test. Data that failed to meet the assumptions of ANOVA were log10 transformed and analyzed using a generalized linear model (GLM) procedure in SAS (version 9.3; SAS Institute Inc., Cary, NC). The significance level was set at α = 0.05, and means were separated using Tukey’s honestly significant difference test. For the transformed data, when ANOVA showed significant differences, the mean separation test was conducted on the transformed data, but nontransformed means were presented for ease in discussion. Interactions between year and treatments were tested and whenever these interactions were significant (p < 0.05), analysis was conducted separately for each year.
Results
Weather at the experimental site
The GDD accumulation from 15 March through harvest was 1908 and 1814 in 2013 and 2014, respectively (Table 1). Precipitation in the same period was 50.0 mm (Figure 1A) and 9.9 mm (Figure 1B) in 2013 and 2014, respectively. The 10-year precipitation average for the study area is 50.5 mm between 15 March and harvest. In 2014, the research site received less than 30% of that. This difference affected the estimated Kc used to calculate irrigation amount and how quickly the canopies filled their allotted space. The estimated crop coefficient (Kc) varied considerably (Figure 2). In 2013, the Kc reached a maximum of 0.84 after 618 GDD. However in 2014, the Kc only reached a maximum of 0.54 at 550 GDD. The Kc affected the estimated vineyard water use as a percentage of the seasonal estimated ETc from budbreak onward.
Phenological progression of Zinfandel 1A in 2013 and 2014 in the southern San Joaquin Valley of California (n = 4).
Seasonal precipitation, estimated mm crop evapotranspiration (ETc), and amount of water applied/vine/week (L) in sustained deficit irrigation (SDI) and regulated deficit irrigation (RDI) treatments in 2013 (A) and 2014 (B) in a commercial Zinfandel vineyard in the San Joaquin Valley of California.
Seasonal crop coefficient (Kc) for a California sprawl canopy on 3.35 m rows as a function of growing degree days (GDD, 10°C base) in 2013 and 2014 in a commercial Zinfandel vineyard in the San Joaquin Valley of California. Each individual Kc was derived from the percent shaded area measured weekly using the equation Kc = 0.017 × percent shaded area.
The Ψl of Zinfandel grapevines responded to irrigation during application in both years (Figure 3). The Ψl was maintained at −1.2 MPa from budbreak through harvest for the SDI treatment. The Ψl was −1.2 MPa from budbreak to fruit set and from veraison to harvest but only −1.4 MPa from fruit set to veraison for the RDI treatment.
Seasonal midday leaf water potential (MPa) response to sustained deficit (SDI) and regulated deficit irrigation (RDI) treatments in 2013 (A) and 2014 (B) in a commercial Zinfandel vineyard in the San Joaquin Valley of California.
Effects of irrigation and rootstock on canopy architecture
There was an effect of year on leaf layer number, leaf area, percent canopy gaps, and number of cluster contacts (Table 2). In 2013, there was an interaction between rootstock and irrigation treatment where a combination of the SDI treatment and Salt Creek produced the greatest PAR transmittance. There were no interactions between rootstock and irrigation treatment in either year for leaf layer number, percent canopy gaps, and number of cluster contacts. Leaf layer number, percent canopy gaps, and PAR transmittance were not affected by rootstock or irrigation treatment. Leaf area was not affected by rootstock in either year. In 2013, SDI increased leaf area by 23% when compared to RDI. However, the same response was not evident in 2014. In both years, the number of cluster contacts was affected by irrigation treatment. In 2013, RDI increased the number of cluster contacts by 21% but decreased it by 19% in 2014.
Effects of rootstock and applied water amount on canopy architecture and microclimate of Zinfandel 1A in the southern San Joaquin Valley of California in 2013 and 2014 (n = 4).
Effect of irrigation and rootstock on yield components
In both years, Zinfandel grafted on Freedom rootstock had greater berry mass, clusters per vine, yield, and skin and berry mass than Salt Creek (Table 3). Likewise, the SDI treatment also had greater berry mass, clusters per vine, yield, and skin and berry mass than the RDI treatment. Cluster weight was only affected in 2014, when RDI decreased it by 10% under SDI. The seed number per berry was consistently higher for Zinfandel grafted on Freedom rootstock than on Salt Creek in both years. There was an effect of experimental year. The general trend in 2014 was a decline in all yield component variables except seed number and weight, which were greater in 2014 than in 2013.
Effect of rootstocks and applied water amounts on yield components of Zinfandel in the southern San Joaquin Valley of California in 2013 and 2014 (n = 4).
Effects of irrigation and rootstock on berry composition
There was no effect of rootstock or irrigation on Brix in either year (Table 4). Juice pH and TA were not affected by rootstock in either year. However, juice pH and TA were consistently affected by irrigation treatment. In both years, RDI increased juice pH but decreased TA at harvest.
Effect of rootstocks and applied water amounts on berry compositions of Zinfandel in the southern San Joaquin Valley of California in 2013 and 2014 (n = 4).
Effects of irrigation and rootstock on anthocyanins
Eleven anthocyanin compounds were identified in the skin of Zinfandel (Table 5). There was a year effect: concentrations of all anthocyanidins increased from 2013 to 2014. The TSA of Zinfandel was consistently affected by irrigation treatment, where SDI treatment increased it over RDI. Also, Zinfandel grafted on Freedom rootstock had greater TSA in the second year of the study than Salt Creek.
Effect of rootstocks and applied water amounts on anthocyanin content (mg/kg) of Zinfandel skins at harvest in 2013 and 2014 in the southern San Joaquin Valley of California.
The concentration of all 3-glucosides increased in 2014 over 2013. Malvidin-3-O-glucoside (m-3-g) was the major anthocyanidin found. In 2013, only m-3-g was affected by rootstock, where it was higher with Freedom rather than Salt Creek. However, in 2014 all anthocyanidins of Zinfandel increased when grafted on Freedom rather than Salt Creek. In both years, SDI increased concentrations of all anthocyanin compounds.
In both years, concentrations of cyanidin, peonidin, and petunidin-3-glucoside-acetate in Zinfandel berry skins were higher when grafted on Freedom than on Salt Creek. Irrigation treatments did not have an effect on the 3-glucoside-acetate anthocyanins, except SDI had more cyanidin-3-glucoside-acetate than RDI in 2013. Rootstocks did not affect the concentrations of coumarates of Zinfandel anthocyanins. However, in 2013 and 2014, SDI increased the concentration of malvidin-and petunidin-3-glucoside-coumarate, respectively.
Acylation of Zinfandel anthocyanins was rarely affected by rootstock or irrigation treatment, although in 2014, vines on Freedom rootstock produced more nonacylated anthocyanins than those on Salt Creek. In contrast, hydroxylation of Zinfandel anthocyanins was consistently affected by rootstock: in both years, vines grafted on Salt Creek had more dihydroxylated anthocyanins and less trihydroxylated anthocyanins than those on Freedom. Irrigation treatments did not affect anthocyanin hydroxylation patterns in either year.
Effects of irrigation and rootstock on leaf area to fruit ratio and water footprint
In both years, leaf area to fruit ratio was affected by irrigation treatment (Table 6). In 2013, the RDI treatment produced 29% less leaf area to fruit ratio, but in 2014, the RDI treatment was 44% greater. In 2014, the leaf area to fruit ratio was greatly affected by the decreased yield, as affected by the RDI treatment after a dry winter.
Effect of rootstock and applied water amount on yield efficiency, total skin anthocyanins produced per hectare, and water productivity of Zinfandel in the southern San Joaquin Valley of California (n = 4).
The water footprint of Zinfandel was affected only by irrigation treatment in 2013. The SDI treatment had a greater water footprint than RDI in 2013. In 2014, rootstock affected the water footprint, which was greatest with the Salt Creek rootstock. As in the previous year, SDI produced a greater water footprint than RDI.
Discussion
Canopy architecture
Generally, there was a lack of rootstock effect on canopy architecture as reported previously (Hatch et al. 2011). However, there was an effect of experimental year where leaf layer numbers increased, while canopy gaps and number of cluster contacts decreased from 2013 to 2014. The decrease in cluster contact numbers was consistent with previous reports of a dry winter hindering canopy development (Chaves et al. 2010, Mendez-Costabel et al. 2014). Leaf area also decreased from 2013 to 2014 as seen in previous studies on rainfall exclusion (Mendez-Costabel et al. 2014). The decreased leaf area in 2014 was attributed to lower water reserves in the soil due to less winter rainfall (Gauthier et al. 2014).
A smaller canopy size and reduced yield due to lower water reserves in soil after a dry winter was reported (Petrie et al. 2004). The dry winter in this study contributed to a deficit irrigation carryover effect in the more stressed RDI treatment in the second year, in which the reduced soil water reserves in the soil resulted in fewer cluster contacts. Across both years, vines grafted onto Freedom rootstock had less PAR transmittance than those on Salt Creek. However, after a dry winter, Zinfandel grafted onto Freedom rootstock produced a canopy that filled its allotted space in the presence of harsh conditions.
Yield components
In 2013, there were few differences in yield components among treatments. However, after a dry winter, rootstock genotype affected berry weight, cluster count at harvest, and subsequently yield: Freedom increased reproductive growth of Zinfandel compared to Salt Creek, which was previously reported as one of the most drought-tolerant rootstocks (Kidman et al. 2014).
The better performance of Zinfandel when grafted onto Freedom rootstock after a dry winter can be attributed to its diverse genetic background, which includes V. vinifera (Garris et al. 2009) and the rootstock’s preference for sandy soils with low fertility (Christensen 2003). Rootstocks with V. vinifera parentage are likely to be more drought-tolerant than American Vitis species due to their geographic origins (Keller et al. 2012). Irrigation treatments had few effects on yield components in 2013. However, after a dry winter, RDI decreased yield below that of SDI. This finding was attributed to a deficit irrigation carryover effect (Lakso et al. 1999, Chaves et al. 2010).
In this study, rootstock genotype not only consistently affected berry and skin mass, but also affected the seed number per berry. Consequently, seed mass per berry was also affected. Freedom rootstock increased seed mass per berry over Salt Creek. An investigation of low-vigor rootstock selections common to central Europe, including parentage crosses of V. berlandieri, V. riparia, V. rupestris, and V. vinifera, found similar differences in berry weights and seed counts per berry (Pulko et al. 2012). That study also reported a correlation between seed count and yield after a dry growing season that corroborated our results, in which Zinfandel grafted on Freedom rootstock had greater yield, seed mass, and seed count per berry after a dry winter.
Conversely, irrigation treatment affected seed mass but not seed number per berry: seed mass was reduced irrespective of seed count. This observation was attributed to the timing of the RDI treatment application. The RDI treatment imposed a greater water stress during a key time of seed development, hindering seed size and weight but not seed count per berry (Keller et al. 2012).
In addition to reported in-season effects of deficit irrigation, there was an impact of drought on yield components. In 2014, the project site was subjected to prolonged drought. Yield components including berry weight, cluster number, cluster weight, and yield per vine declined due to lack of winter precipitation, as consistent with other reports (Mendez-Costabel et al. 2014). In contrast, the seed weight per berry increased from 2013 to 2014, which can be attributed to preferential resource partitioning under water stress (Gauthier et al. 2014).
Berry composition
In general, there were limited effects of rootstock on berry composition, as confirmed by other reports (Pulko et al. 2012, Stevens et al. 2008). The effects of irrigation regimes on berry composition have been well documented and are generally consistent with our results. Compared to SDI, the more stressed RDI treatment decreased TA and increased pH, as reported previously (Keller et al. 2012).
Anthocyanin concentration and composition
The growing region’s temperature and the amount of light subtending the clusters can affect anthocyanin biosynthesis, accumulation, and degradation (Bergqvist et al. 2001, Spayd et al. 2002). In 2013, there were few differences in anthocyanin and TSA concentration associated with rootstock genotype. These results closely paralleled the absence of a rootstock genotype effect on anthocyanin concentration reported previously (Berdeja et al. 2014). In 2014, a warmer year than 2013, Zinfandel grafted on Salt Creek was slower to fill its allotted canopy space, and clusters were more directly exposed to sunlight early in the growing season. This provided evidence for decreased m-3-g and TSA concentration in Zinfandel grafted on Salt Creek rather than Freedom. It is generally accepted that, as fruit exposure to sunlight increases, berry composition improves (Bergqvist et al. 2001). However, a concomitant increase in berry temperature with greater exposure to direct sunlight can decrease berry skin anthocyanin concentration in hot climates (Bergqvist et al. 2001). The increased cluster temperature and the region’s high GDD accumulation decreased anthocyanin concentration below that of grapevines grafted on Freedom that filled allotted canopy space earlier, as reported elsewhere (Wessner and Kurtural 2013).
The composition of anthocyanin acetyl glucosides was consistently affected by rootstock genotype. There is little information on the effect of rootstock genotype on the anthocyanin acetyl glucoside concentration in the literature. However, there is agreement that greater cluster shading results in a shift toward acetylated anthocyanins (Downey et al. 2004). The concentrations of cyanidin and petunidin-3-glucoside acetate were consistently greater when Zinfandel was grafted on Freedom instead of Salt Creek. However, the proportion of acylated versus nonacylated anthocyanins was not affected by rootstock genotype in either year.
There is broad agreement that higher temperature, rather than light, results in a shift toward coumaroylated anthocyanins (Downey et al. 2004, Spayd et al. 2002, Cook et al. 2015). Although we saw no shift towards coumaroylated anthocyanins in 2014, a warmer year than 2013, the petunidin-3-glucoside-coumarate concentration increased after a dry winter, as previously reported (Tarara et al. 2008). However, the malvidin-3-glucoside-coumarate concentration decreased from 2013 to 2014. As with anthocyanin-glucoside-acetates, further hot climate research into the mechanisms controlling anthocyanin derivatives is needed.
Experimental year affected anthocyanin concentration. The combination of prolonged drought, limited canopy development, and decreased water application created conditions similar to those reported with water stress. The consistent production of a higher proportion of trihydroxylated anthocyanins by vines grafted onto Freedom would enable growers to produce more stable anthocyanin concentrations in berry skin in hot climates (Cook et al. 2015). Freedom rootstock increased the proportion of malvidin, petunidin, and delphidin glucosides relative to cyanidin and peonidin glucosides (Castellarin et al. 2007, Cook et al. 2015). Additionally, as the relative skin tissue mass increased, the increased anthocyanin concentration associated with vines grafted on Freedom rootstock was consistent with earlier reports (Diago et al. 2012, Cook et al. 2015). However, these reports were not consistent with in-season results: berries grown on Salt Creek had more skin tissue mass but lower anthocyanin concentrations than those on Freedom. The influence of heat, due to reduced canopy fill early in the season, overcame the increase in skin mass in determining anthocyanin concentration. Freedom rootstock filled its allotted canopy space in spite of drought-induced water stress that increased the skin tissue mass. Freedom rootstock increased total anthocyanin concentration over Salt Creek across a three-year study (Rizk-Alla et al. 2011).
The SDI treatment consistently had more m-3-g and total skin anthocyanins in the berry skin than RDI. This was in contrast to previous work, where application of irrigation deficits between grape berry growth stages I and II, similar to the RDI treatment presented here, resulted in greater anthocyanin accumulation (Kennedy et al. 2002, Castellarin et al. 2007, Terry and Kurtural 2011). The results presented here provide evidence that the amount of water applied to the RDI treatment was less than the grapevine needed. Similar results were reported in a hot climate with Merlot grapevine, where reducing applied water amounts with a similar RDI treatment provided no further benefit to anthocyanin composition over SDI (Cook et al. 2015). Irrigation treatments had few effects on accumulation of dihydroxylated, trihydroxylated, or percentage of acylated anthocyanins, in contrast to a recent study conducted in the northern SJV (Cook et al. 2015).
Leaf area to fruit ratio and water footprint
In general, rootstock genotype had little affect on leaf area to fruit ratio. Traditionally, the leaf area to fruit ratio required to optimize maturity at harvest ranged from 0.8 to 1.2 m2/kg for single-plane, non-shoot-positioned canopies (Terry and Kurtural 2011). This trial was successful at maturing Zinfandel on a mechanically pruned system, regardless of rootstock selection, to a commercial standard of ~20% TSS (Wolpert 1996). The 0.6 to 0.9 m2/kg leaf area to fruit ratio in this study is below this traditional standard. The leaf area to fruit ratio of Zinfandel was not consistently affected by applied water amount. In 2013, the SDI treatment had greater leaf area to fruit ratio than RDI, owing to greater leaf area per vine with similar yield. Conversely in 2014, the leaf area to fruit ratio of RDI was greater, owing to similar leaf area per vine but lower yield. These results suggest a threshold of >0.60 m2/kg for single-plane, non-shoot-positioned canopies is sufficient for Zinfandel to reach commercial maturity in the southern SJV.
The water footprint of viticulture has drawn attention from researchers, as once-fertile growing regions such as the SJV continue to battle anthropogenic climate change and water shortages (Chaves et al. 2010). In agreement with Williams (2012), this project reported increased water crop productivity and subsequent decrease in water footprint (Williams 2014), with limited reductions in yields for RDI treatments. Furthermore, decreased yield associated with the more stressed RDI treatment occurred only after a dry winter (Chaves et al. 2010). In this study, water productivity increased as yield increased or as applied water amounts decreased. Water footprint decreased across all treatments from 2013 to 2014, regardless of the decreased yield associated with drought. This decrease in water footprint was an effect of less water applied due to reductions in canopy development. There were no differences between rootstock water footprints in 2013. After a dry winter, scion grafted on Freedom rootstock exhibited decreased water footprint. This observation can contribute to the sustainability and drought tolerance of Freedom rootstocks as access to irrigation water continues to limit growers in hot climate regions. In both years, the RDI treatment was more efficient at using applied water resources. However, in 2014, the RDI treatment decreased yield. This decrease in yield was commercially unacceptable and cannot be recommended, regardless of amelioration of water footprint. Our results suggest that SDI combined with Freedom rootstock can enhance water productivity within the southern SJV through limited reductions in yield, higher anthocyanin concentrations in berry skin, and increased water productivity in a resource-limited area.
Conclusions
The interactive effects of rootstock and applied water amounts on canopy architecture, berry development, yield components, yield efficiency, anthocyanin concentrations, and water footprint within a hot climate were measured. As anthropogenic climate change continues to affect agriculture, growers and winemakers must deal with untoward weather conditions, as in the severe drought affecting this study. During a prolonged drought, Zinfandel grafted on Freedom rootstock was successful at limiting reductions in yield. Additionally, Freedom increased accumulation of TSA better than grapevines grafted on Salt Creek rootstock and decreased the water footprint, which is of importance as California deals with increased competition for water resources. As in previous studies, irrigating at 50% ETc from fruit set to veraison decreased the water footprint in both years. However, the increased water stress decreased yield and TSA. Due to the region’s reliance on high yield, the reduction in water footprint was not large enough to outweigh the loss in yield. A combination of Freedom rootstock and irrigating to 80% ETc is recommended for growers in the hot climate of the southern SJV based on yield, berry composition, and reduction of water footprint.
Acknowledgments
This study was funded in part by the American Vineyard Foundation and Bronco Wine Company Research Chair Trust Funds. A graduate student stipend was provided by the American Vineyard Foundation to Clinton C. Nelson. The authors also acknowledge West Coast Grape Farming Inc., Faustino Valdez, Geoffrey Dervishian, Michael Cook, Andres Mendez, Kerry Fitzgerald, Yaritza Aguirre, and Tiffany Gündüz for technical assistance during the execution of this trial. The authors also acknowledge the critical review of the manuscript by Dr. Matthew Fidelibus.
- Received May 2015.
- Revision received August 2015.
- Accepted August 2015.
- Published online December 1969
- ©2016 by the American Society for Enology and Viticulture
Literature Cited
Vol 67 Issue 1
Jump to section
Related Articles
Cited By...
More from this TOC section
Similar Articles
