Regulated Deficit Irrigation Effect on Yield and Wine Color of Cabernet Sauvignon in Central California

Regulated deficit irrigation (RDI) is a management strategy that on grape can improve shoot/fruit ratio, water efficiency, and wine quality but has the potential to reduce yield. As part of a study on the influence of RDI on leafhopper density, we evaluated the effects on grape yield, berry size, berry soluble solids, and wine color. The studies were conducted at commercial vineyards in the San Joaquin Valley and in the Paso Robles region, CA, with Cabernet Sauvignon as the cultivar. Water deficits were imposed at either 50% (moderate deficit) or 25% (severe deficit) of standard irrigation (the control) for a period of 3 or 6 weeks and initiated at berry set, leafhopper egg hatch, or veraison. Deficit irrigation decreased berry weight by 16.1% at the San Joaquin Valley site (Aliso) and 11.7% at one of the Paso Robles sites (Frankel) but did not differ at the other site (Steinbeck). Yield was decreased by the deficits by 18.1% at Aliso, 26.7% at Frankel 2001 (but not 2002), and 24% at Steinbeck. Wine color density was increased by 21.8% at Aliso, 34.4% at Frankel 2001 (but not 2002), and did not differ at Steinbeck. Soluble solids did not differ among treatments at any site. There was no difference in berry weight, yield, or color between the moderate and severe deficits. It appears that in central California, RDI such as these are likely to reduce yield but are only one factor among many variables affecting quality such as wine color. In perennial fruit and nut cropping syssome resulted in yield decreases (Egea et al., tems, water can be applied at a reduced rate 2010; Iniesta et al., 2009), and others found during a defined phenological period while yield increases (Chalmers et al., 1986; Mitchell maintaining standard irrigation during the et al., 1984). rest of the season (Chalmers et al., 1981). On wine grapes, RDI is distinguished from This has come to be known as RDI and has partial root zone drying, in which the deficit received a great deal of attention and research variable is not one of time, but of space, as focus. The potential advantages of RDI are applied water is alternated from one side of in improvements in shoot-to-fruit load ratios the vine to the other (Dry and Loveys, 1998). (Iniesta et al., 2009; Mitchell et al., 1984), The timing of RDI on grapes has typically inwater efficiency or productivity (Cui et al., volved imposing the deficit from the grapevine 2009; Egea et al., 2010; Garcia-Tejero et al., phenological periods of berry set to veraison 2010), and fruit quality (Garcia-Tejero et al., or veraison to ripeness (harvest) (Matthews 2010; Papenfuss and Black, 2010). The main and Anderson, 1988; Matthews et al., 1987), potential disadvantage is on fruit production, berry set to harvest (Acevado-Opazo et al., although results of RDI on yield have been 2010; Chaves et al., 2007; dos-Santos et al., mixed. Some studies showed a neutral effect 2007; Shellie, 2006), and, less often, bud­ (Cui et al., 2009; Papenfuss and Black, 2010), break to bloom (Goodwin and Jerle, 1989) or budbreak to veraison (Hamman and Dami, 2000). Grape berry development occurs in three phases: post-berry set to veraison (rapid Received for publication 9 Apr. 2012. Accepted for cell division followed by cell enlargement), a publication 11 June 2012. lag phase (cessation of cell enlargement and Thanks to cooperators Martin Britz, Kip Green, and initiation of sugar accumulation), and, fiGurmit Singh of Britz Farming (Aliso Vineyard); nally, veraison to ripeness (cell enlargement Warren Frankel at Frankel Vineyards; and Tim Lindquist and Howie Steinbeck at Steinbeck and initiation of anthocyanin accumulation) Vineyards. Field and laboratory assistance was (DeLuc et al., 2007). Which phenological provided by students Catherine Albers, Christopher period the deficit is imposed in should have Boisseranc, Paul Frankel, Bryce Gross, Jo Harper, a distinct effect. Matthews and Anderson Alfredo Koch, Rachel Power, Kevin Robertson, (1987) and Matthews et al. (1988) found that and Dory Wachtel. We are grateful to the the deficits imposed pre-veraison or post-veraison California Agricultural Technology Institute (CATI) increased anthocyanins and other phenolics, at California State University, Fresno, the California produced smaller berries and reduced yield, State University Agricultural Research Initiative but the effects on berry size and yield were (ARI), and the Viticulture Consortium for funding these studies. more pronounced with pre-veraison com­ To whom reprint requests should be addressed; pared with post-veraison deficit. McCarthy e-mail [email protected]. (1997) found that the greatest reduction on yield was when deficit was initiated just after bloom but before berry set and compared with the standard irrigation, pre-veraison and post-veraison lowered berry size, but there was little difference between these two pe­ riods. (Acevado-Opazo et al., 2010) found that RDI imposed between berry set and veraison or harvest increased concentration of anthocyanins and decreased berry diame­ ter (but not berry weight), but only with a deficit intensity of 26% or less but not that of ;50% of the standard irrigation. On cultivated grape in California, it is common practice to irrigate at ;100% of evapotranspiration (1.0 ETc) or more from 4 to 6 weeks after budbreak until just be­ fore harvest. However, Williams et al. (2010) showed that on Thompson Seedless, seasonwide irrigation at 0.80 ETc can take place with no significant yield loss. To evaluate the impact of RDI on leafhopper density (pri­ marily Erythroneura elegantula Osborn), we imposed RDI of either 50% or 25% of standard irrigation during second-generation leafhopper development, which about corre­ sponds to the period between berry set and veraison. Erythroneura spp. overwinter as adults and begin feeding on grape tissue shortly after budbreak in the spring (late March to early April) (Costello, 2011). After mating, females lay eggs within the leaves. The first-generation nymphs hatch, develop through five stages (instars), and in the Paso Robles region of California, molt to adult­ hood by early to mid-June. Second-generation nymphal hatch is typically midto late July. Erythroneura spp. are sensitive to vine water status (Daane and Williams, 2003), and we have shown that RDI during second-generation nymphal development resulted in a signifi­ cant decrease in leafhopper nymphal density (Costello, 2008; Costello and Veysey, 2012). We present the results of these studies on fruit production and wine color as an indicator of quality. Water application and stomatal con­ ductance (gS) results from one of the study sites and years (Frankel in 2001) are pre­ sented here; otherwise, these data can be found in Costello (2008) and Costello and Veysey (2012). Materials and Methods Details of experimental design and cul­ tural practices for the Aliso and Frankel vineyards are in Costello (2008) and for the Steinbeck vineyard in Costello and Veysey (2012), but a summary is presented here. The study sites were mature, commercial vine­ yards located in central California. The Aliso vineyard was located in Madera County with a mean annual rainfall of 204 mm and mid­ summer (July through August) mean high and low temperatures 35.5 and 16.3 °C, re­ spectively. The Frankel and Steinbeck vine­ yards were located in San Luis Obispo County with mean annual rainfall 373 mm and mid­ summer mean high and temperatures 33.4 and 10.8 °C, respectively. The cultivar at all of the sites was ‘Cabernet Sauvignon’ with a cordon-trained, spur-pruned system but in HORTSCIENCE VOL. 47(10) OCTOBER 2012 1520 Winter 2001–02, the vines at the Frankel site were retrained to a head-trained, cane-pruned system. At Steinbeck, the study plots were moved from the southwest corner of the vineyard (2002) to the northwest (2003), where soil type was less variable. Each experiment was designed as a ran­ domized complete block with treatments repli­ cated four times. At the Frankel and Steinbeck sites, the deficit treatments were undertaken to reduce the control irrigation to 50% or 25% of standard irrigation, which was close to 1.0 ETc; at Aliso, only the 50% deficit was undertaken. These will hereafter be referred to as moderate (50%) or severe (25%) deficit. At the Steinbeck site, an additional split plot treatment varying the duration of the deficit (3 weeks vs. 6 weeks) was included. Deficit irrigation typically was initiated at berry set and maintained until veraison, although at the Aliso site, two variations of this period were included: leafhopper egg hatch (eclosion) to veraison and veraison to harvest. Deficit treat­ ments were compared with a control based on the standard irrigation rate set by the vineyard manager at each site, who was assumed to be irrigating 0.8 to 1.0 ETc throughout the season. The control irrigation estimation for the period of deficit was 0.89 ETc at Aliso and 0.92 ETc at Frankel in 2002 (Costello, 2008) and at Steinbeck 0.92 ETc and 0.94 ETc in 2002 and 2003, respectively (Costello and Veysey, 2012). Estimated control and deficit treatment water applied at Frankel in 2001 as in Table 1. At each site, deficits were induced for a period of ;6 weeks, and at Steinbeck, a treat­ ment was added to compare 3-week duration with the 6-week duration. At Aliso, three deficit initiation treatments were used: berry set (1 June), leafhopper egg hatch (eclosion) (29 June), and veraison (20 July), and each was imposed for 6 weeks. At Frankel in 2001, the deficits were initiated on 28 June and im­ posed until 9 Aug.; at Frankel in 2002, the deficits were initiated on 20 June until 1 Aug. At Steinbeck, deficit initiation took place on 21 June in 2002 and 2003 and imposed until either 12 July (3-week deficit) or 2 Aug. (6-week deficit). Before the imposition of the deficit irri­ gation treatments, all vines in the study area were watered according to each grower’s irrigation schedule. After the deficit, irriga­ tion was set to 80% of the grower standard at Aliso and to 100% of the grower standard at Frankel and Steinbeck. Stomatal conductance (Frankel in 2001) was measured with an LI-6200 CO2 poro­ meter (LI-COR, Lincoln, NE). Leaves se­ lected for measurement were mature and in full sun. Grapes were harvested on 17 Sept. at Aliso, 30 Sept. at Frankel 2001, 1 Oct. at Frankel 2002, 8 Oct. at Steinbeck 2002, and 3 Oct. at Steinbeck 2003. Fruit was harvested and weighed from four vines per plot at Aliso or four cordons per plot or subplot at Frankel or Steinbeck. We took berry samples (;100 berries per plot at Aliso, 50 per plot at Frankel, and 80 per plot at Steinbeck) for estimates of berry sugar and size. Berry sugar was measured as soluble solids (°Brix) with a temperature-compensating refractometer (Leica®, Buffalo Grove, IL), and berries were weighed en total and then divided by the number of berries for estimated weight per berry. Wine was made by processing the grapes through a stemmer/crusher, adding potassium metabisulfite and inoculating with yeast, fer­ menting in open-topped vessels covered with cheesecloth, and pressed when degrees ball­ ing were close to zero. The must was then pressed in a basket press and allowed to finish in a glass carboy. The wine was racked, and samples were analyzed for color using a spec­ trophotometer (Ivyland, PA). Most data were log-transformed, except for color, which was transformed by reflect and inverse transformation. All data were analyzed by analysis of variance (ANOVA) with mean separation by Tukey’s honestly significant difference (SAS Institute, 2010). Differences were considered statistically sig­ nificant at P < 0.05. For the Frankel yield data, there was a significant year*treatment interaction (F = 11.93, df = 2, 42, P < 0.001), so each year’s data were analyzed separately.