Abstract
Three vineyard floor management strategies—reflective (white) geotextile mulch, black geotextile mulch, and an herbicide strip in vine rows—were evaluated with regard to canopy light and temperature, vine growth, and fruit composition of Cabernet franc. Using a randomized complete block design, two trials were conducted at commercial vineyards in the Finger Lakes Region of New York during 2004 to 2005. Black and white geotextiles in 2-m-wide or 5-m-wide strips almost doubled vine size after two years, but did not affect overwinter primary bud survival compared with mowed sod row middles with undervine herbicide control treatment. In both experiments, the white geotextile mulch significantly increased sunlight reflected from the vineyard floor into the vine cluster zone, especially early in the growing season. Vines mulched with white geotextile had greater yields, but there were no significant differences in fruit ripening time or fruit chemical composition at harvest, and minimal differences in vine nutrient status among the floor management systems. The increased yield values in reflective geotextile plots were not sufficient to compensate for the substantially greater costs of geotextiles relative to the standard practices of mowed sod row middles with undervine herbicide weed control.
Growers in cool, humid regions face many challenges with Vitis vinifera winegrapes, particularly red varieties such as Cabernet Sauvignon, Cabernet franc, and Pinot noir. Common problems include delayed veraison, insufficient ripening of grapes, and inadequate primary bud survival during cold winters. During cool, wet, cloudy summers such as those of 2003 and 2004 in upstate New York, vineyards often suffer reduced yields, delayed berry ripening, and poor flavor development. The potential for winter cold injury to buds is also exacerbated.
Increasing fruit exposure to sunlight can increase soluble solids and reduce acidity in grapes (Kliewer et al. 1988) and also increase anthocyanins and total polyphenols (Jackson and Lombard 1993, Smart et al. 1990). Excessive shading of grapes has been shown to delay ripening (Smart et al. 1988) and increase Botrytis infection and herbaceous character in wines (Smart et al. 1990). Shoot positioning, shoot thinning, and leaf removal are common practices in cool-climate regions to improve canopy microclimate and fruit exposure to sunlight. Reflective mulches or soil covers might also increase sunlight in the fruit zone and consequently improve berry ripening and flavor development. In contrast to shoot thinning and leaf removal, reflective mulches would not reduce the total photosynthetic capacity of vines and could modify the canopy microclimate to improve ripening and disease suppression.
Reflective mulches have reportedly improved red coloration, advanced ripening, and increased fruit size in apple (Andris and Crisosto 1996) and sweet cherry (Hansen 2005). Studies in New Zealand have suggested that reflective mulches may advance veraison and improve wine quality in Cabernet Sauvignon and Pinot noir (G. Creasy and A. Pickering, unpublished data). There are only a few published reports on the effects of reflective mulches or geotextiles (open-weave, durable fabrics that block sunlight and weed growth but permit water infiltration and gas exchange) on fruit composition or yield in established vineyards (Vanden Heuval and Neto 2006, Pearson 2004).
The purpose of this study was to determine if white or black geotextile mulches could provide useful alternatives to conventional vineyard floor management by affecting canopy microclimate, time of fruit ripening, incidence of bunch rot, fruit composition, yield, vine size, and overwinter primary bud survival. We evaluated three vineyard floor management systems: nonreflective (black) geotextile, reflective (white) geotextile, and the conventional system of mowed sod row middles and undervine herbicide strip.
Materials and Methods
Experimental design.
Test plots were established in 2004 at Sheldrake Point Vineyard (SPV) at 135 m elevation near Cayuga Lake, and at Wagner Vineyard (WV) at 270 m elevation near Seneca Lake in upstate New York. The two vineyards were about 10 km apart and had similar weather patterns during this study. The soil at both vineyards is a Honeoye silt loam (fineloamy, mixed, mesic, Glossic Hapludalf ) with soil pH and organic matter content averaging about 6.8 and 4.5%, respectively. The SPV site slopes toward the southeast and the WV site slopes toward the southwest. The SPV site was a dairy farm until 10 years ago and decades of manure applications elevated soil nitrogen, causing some problems with excessive vine vigor. The WV site has been a vineyard for over 30 years; vine size is usually smaller than at SPV, although excess vigor has been a problem in summers with abundant rainfall.
The experiment at SPV consisted of 108 seven-year-old Cabernet franc vines on Couderc 3309 rootstock, trained to a Scott Henry system. The experiment at WV consisted of 90 nine-year-old Cabernet franc vines on S04 rootstock, also trained to a Scott–Henry system. Each experiment was set up in a randomized complete block design, with six treatment replicates at SPV and five at WV. Each block was assigned to a different row and the vineyard floor management system (VFMS) treatments were randomized within each row. The experimental plots were chosen carefully among several thousand vines at each planting, selecting groups of six contiguous vines that were relatively uniform in size and were pruned to four 12-node canes per vine. To minimize edge effects, only the four middle vines in each plot were used for sample measurements. Extreme cold in January 2004 and February 2005 damaged some primary buds at both vineyards, causing fruitful bud numbers to vary somewhat from our target range.
The experimental treatments at SPV included white geotextile mulch (Dewitt Industries, Sikeston, MO) in a 5-m-wide strip on both sides of the vine row (i.e., row-to-row, covering the entire vineyard floor), black geotextile mulch in a similar 5-m-wide strip, and the grower’s usual system of mowed grass row middles with a 0.5-m paraquat herbicide strip under the vine rows as the control treatment. Geotextiles were installed in late May and removed in late October both years, and glyphosate had to be applied before installing the white geotextile at SPV to suppress the underlying vegetation. Geotextiles were secured to the ground with large metal staples along all edges.
The treatments at WV included white geotextile mulch in 2-m-wide strips centered on the vine rows, black geotextile mulch in a similar 2-m-wide strip, and the grower’s usual system (mowed grass row middles with a 2-m-wide preemergence herbicide strip under the vines) as the control treatment. Geotextiles at WV were installed in May 2004 and remained in place until the end of 2005; they were secured with large metal staples along the vines and their outer edges were buried with a thin berm of soil at the row middles.
Light and temperature measurements.
Light and temperature were measured at 1.25 m above the ground from June through October in 2004 and 2005. Sunlight in the cluster zone was measured every two weeks between 1000 and 1400 hr with an LI191 line quantum sensor and LI250A light meter (LICOR, Lincoln, NE) as both overhead light (meter pointed upward above clusters) and reflected light (meter pointed down toward the vineyard floor below clusters). In 2004, temperatures in the cluster zone (about 1.25 m above the ground) in each plot were continuously recorded with miniature data loggers (Kooltrak, Palm Beach Gardens, FL) placed in plastic bags and covered with folded reflective cardboard. In 2005, HOBO Pendant dataloggers (Onset, Pocasset, MA) were used to continuously record both light and temperature in the cluster zone at SPV, while Kooltrak temperature loggers were used again at WV.
Vine size, bud survival, and elemental nutrient analysis.
Vegetative growth of vines was determined for the previous growing season by weighing dormant cane prunings during the subsequent winter. In dormant pruning both experiments, the four best canes (totaling 40 to 50 nodes) were retained on each vine for the subsequent crop. All cane prunings from the previous growing season were grouped for each vine, and their weights were recorded to estimate vine size. Overwinter survival of primary buds was measured by counting the percentage of nodes with fruitful primary shoots emerging on each vine after budbreak in May.
Vine nutritional status was measured by elemental petiole analysis (Shaulis and Kimball 1956). Twenty petioles were collected from each plot at bloom (late June) for nitrogen (N) analysis, and at veraison in mid-August for all other essential nutrients. The youngest mature leaf from a primary bearing shoot was selected and its leaf blade was removed. Petioles were washed with a mild detergent, rinsed, and dried at 25°C to constant weight. Nutrient contents were determined by standard methods at the Cornell Nutrient Analysis Laboratory (Ithaca, NY). Total carbon (C) and N were measured by Dumas combustion (Horneck and Miller 1998) using an automatic elemental analyzer (ThermoQuest Italia, Milan, Italy).
Berry sampling and juice extraction.
At the outset of veraison (early to mid-August) each year, 100-berry samples were randomly collected from four vines in each plot every 7 to 10 days until harvest, when 200-berry samples were collected. In 2004, samples were weighed immediately after collection and frozen at −20 °C until analysis; in 2005, berries were processed immediately after harvest.
Frozen berries were thawed at room temperature, heated in a water bath to 80°C to bring tartrates into solution, and then cooled to 20°C. Berries were then pressed through cheesecloth with a conical press and wooden pestle. Sugar content of the juice was measured as soluble solids using a Pocket PAL-1 digital refractometer (Atago, Bellevue, WA). Juice pH was measured with a Symphony SB20 pH meter (VWR, West Chester, PA). Titratable acidity was measured with a DL12 automatic titrator (Mettler-Toledo, Columbus, OH) using 0.1 N NaOH and an endpoint of pH 8.2.
For berry skin analysis, 30berry subsamples were thawed at 25°C. Pulp was squeezed from the berries and any remaining pulp was removed by hand. The skins were rinsed with distilled water, blotted dry with a paper towel, and weighed. Skins were extracted in acidified methanol (1% HCl v/v) as described (Metevier et al. 1980). Grape skins were placed in a 250-mL Erlenmeyer flask, and 22 mL acidified methanol was added for each gram of skins. The flask was covered with parafilm and left in the dark at 20°C for 48 hr. The extract was then decanted and centrifuged at 4°C and 10,000 rpm for 5 min. The supernatant was decanted into a 20-mL plastic bottle and stored at −20 °C until analysis.
For whole berry analysis, 30-berry subsamples were taken from the 100-berry samples before harvest, and 60-berry subsamples were taken from 200-berry samples at final harvest. Berries were cut in half and stored at −20 °C until they were freezedried. Freezedried berries were ground with mortar and pestle, and extracted as described (Kim and Lee 2002). One gram of freeze-dried sample was combined with 10 mL 80% MeOH in a 50-mL centrifuge tube with the headspace filled with nitrogen gas, and placed on a shaker table for 10 min at 80 rpm. Each sample was then centrifuged at 4°C at 10,000 rpm for 20 min. The extract was decanted into a 25-mL graduated cylinder and each sample was then extracted with another 10 mL 80% MeOH as before. MeOH was added to the graduated cylinder to bring it to a volume of 25 mL. The extract was then transferred to a 50-mL vial with the headspace filled with N gas and stored at −20°C until analysis.
Harvest.
At SPV, fruit was harvested on 13 Oct 2004 and 17 Oct 2005. Fruit from WV was harvested 12 Oct 2004 and 19 Oct 2005. The number of clusters and total yield were recorded for each vine. The incidence of cluster rot was also assessed at SPV.
Must analysis.
In 2004, fruit from the final harvest at SPV was processed into wine. Fruit from each fourvine plot was crushed separately and a 100-mL sample of must was collected for prefermentation measurements of soluble solids, pH, and titratable acidity with an automated wine analyzer (FOSS North America, Eden Prairie, MN). All must from replicates within each treatment was then combined for fermentation in 20-L carboys, and the three resultant wines were chemically analyzed using the methods described above for berry samples.
Analysis of anthocyanins, total phenolics, and antioxidants.
Berry and skin extracts were prepared as described previously. Juice and must samples were centrifuged at 4°C and 15,000 rpm for 10 min prior to analyses. Anthocyanin was measured using the pH differential method (Giusti and Wrolstad 2001). Samples were diluted to an absorbance range of 0.1 to 1.0 and all measurements were performed in triplicate. A pH 1.0 buffer was prepared with potassium chloride and a pH 4.5 buffer was prepared with sodium acetate. Each sample was diluted with each buffer in a test tube, shaken, and allowed to stand 30 min. The absorbance was measured at 520 and 710 nm using distilled water as a reference blank. Total anthocyanin was calculated as malvidin-3-glucoside equivalents using 529 as the molecular weight and 28,000 as the molar absorptivity.
Total phenolics were determined as described (Singleton and Rossi 1965). Samples were diluted to an absorbance range of 0.1 to 1.0 and all measurements were performed in triplicate with a blank. A 0.2-mL aliquot of sample was added to 2.6 mL deionized distilled water in a test tube. Then 0.2 mL of Folin and Ciocalteu phenol reagent was added and the mixture was shaken. After 6 min, 2 mL 7% Na2CO3 was added and the mixture was shaken again. After incubation for 90 min at room temperature, the absorbance at 750 nm was measured against a blank. Total phenolics was calculated using a standard curve for gallic acid. Values were expressed as mg gallic acid equivalents (GAE) per g fresh berry sample, or mg GAE per liter must.
Vitamin C equivalent antioxidant capacity (VCEAC) was measured as described (Kim et al. 2002). A phosphate buffer solution containing 100 mM K2HPO4 and 150 mM NaCl was prepared and stored at 4°C until used. For each day of measurements, a new radical stock solution was prepared by first heating the phosphate buffer solution (PBS) in a water bath to 68°C. Then 1.0 mM AAPH (2,2′-azobis[ 2-amidinopropane] HCl), a radical initiator, and 2.5 mM ABTS (2,2′-azinobis[ 3-ethylbenzthiazoline-6-sulfonic acid] diammonium salt) were added to the PBS and the mixture was heated at 70°C for at least 20 min, agitating every 5 min. The radical solution was allowed to stand at room temperature for 10 min and filtered through an Acrodisc LC13 PVDF 0.45-μm filter. The filtered solution was diluted with distilled water so its absorbance at 734 nm was 0.65 ± 0.02. The diluted radical solution was incubated at 37°C during measurements. Samples were diluted to an absorbance range of 0.1 to 1.0 and each set was compared with a blank. Twenty μL sample was mixed with 980 μL radical solution and incubated for 10 min at 37°C. The absorbance was measured at 734 nm. VCEAC was calculated as the reduction in absorbance at 734 nm over 10 min, compared to a standard absorbance curve for ascorbic acid, and was expressed as mg VCEAC per g fresh sample.
Economic analysis.
Costs for each treatment, excluding hilling up soil over graft unions, were calculated on a perhectare basis prorated over a three-year period, assuming that the geotextiles would be usable for at least three years. Data from the second year were used to estimate costs for the third year. Equipment costs and depreciation were not included in cost estimates, and mowing and herbicide costs were based on published values (White 2005). All costs were based on a planting density of 1994 vines per hectare, planted at vine by row spacings of 1.8 by 2.7 m. The relative cost/benefit ratios of treatments were based upon observed differences in yield compared to the control treatment, assuming a market value of $1654 per ton for Cabernet franc grapes.
Statistical analysis.
Statistical analysis was done with SAS software (SAS Institute, Cary, NC). Because treatment variances were unequal in reflected light measurements, the Kruskal-Wallis test (proc npar1way) was used to assess treatment differences. Mixed model procedure (proc mixed) was used to determine significant differences between treatments at p = 0.05 for all other measurements, and blocks were treated as random effects. Means separation was with Tukey-Kramer procedure for all data, unless otherwise noted. Regression procedure (proc reg) was used to analyze components of yield. Graphs were created with SigmaPlot (Systat, San Jose, CA).
Results
Light and temperature measurements.
At SPV, the row-to-row white geotextile reflected significantly more sunlight off the vineyard floor than the black geotextile or control treatments, with the greatest treatment differences from early to midsummer (Figure 1⇓). Between veraison and harvest, variations in reflectance were attributed mostly to varying weather conditions: more sunlight was reflected on sampling dates without cloud cover. In 2004 at SPV, there were significant differences in reflected sunlight between the white geotextile and the other two treatments on all sampling dates except 20 Aug and 16 Oct. In 2005, there were significant differences in reflected light between the white geotextile and the other two treatments for all sampling dates except 28 Oct.
From early to midsummer at the WV experiment, the narrower white geotextile treatment also reflected significantly more sunlight upward from the vineyard floor than the black geotextile or control treatments (Figure 2⇓). For the remainder of the season, the considerable day-to-day variation in reflectance was due to weather conditions; more light was reflected on sunny sampling dates than cloudy dates in both years. Ambient temperatures within the vine canopy at 1.2m aboveground averaged 18.1°C during the 2004 growing season (a cool, wet year), and 21.6°C during 2005 (a hot, dry year), but there were no significant treatment effects on canopy temperatures at either vineyard during this study (data not shown).
Vine size, bud survival, and nutrient analysis.
Vine size did not differ among treatments at WV during the first year, but after the second growing season pruning weights were nearly doubled in white and black geotextile plots compared with the control treatments at both vineyards (Table 1⇓). Overwinter survival of primary buds was greater in 2005 at SPV and greater in 2006 at WV, but was not significantly different among treatments at either vineyard during two years with extensive winter cold injury to grapevines in the Finger Lakes Region.
There were few significant differences in petiole nutrient content of vines among the treatments at either vineyard (Table 2⇓). At SPV in 2005, petiole N concentrations were higher in black and white geotextile plots than in the control. In both years, petiole Cu levels were also higher in the two geotextiles versus the control treatment. Petiole N concentrations were numerically higher in both geotextiles versus the control at WV in 2005 (data not shown), but these differences were not statistically significant.
Fruit composition.
Soluble solids concentrations were higher in control berries than for either geotextile treatment during the latter part of ripening in 2004 at SPV (Table 3⇓). There were negligible differences in juice soluble solids, pH, titratable acidity, or anthocyanins at SPV during ripening in 2005 (data not shown), but there were differences in fruit composition at harvest. Fruit from black geotextile plots had higher titratable acidity and less anthocyanin than the control treatment. Dry matter content of fruit was lower (and water content higher) when grown in black or white geotextile plots.
Fruit in black geotextile plots at SPV generally had lower total phenolic concentrations and antioxidant activity than the other two treatments during both years (Table 4⇓). In 2004, fruit in the black geotextile plots had lower total phenolics and antioxidant activity than fruit in white geotextile plots. Similar trends were evident in 2005, when fruit from black geotextile plots had lower total phenolics compared to white geotextile and control treatments, and lower antioxidant activity compared to fruit in white geotextile plots.
There were no significant differences in fruit soluble solids, pH, anthocyanin, total phenolics, or average berry weight among the treatments at WV during the 2004 season (data not shown). On 10 and 28 Sept 2004, the fruit from black geotextile plots at WV had higher titratable acidity than the control, and this trend continued through harvest (Table 5⇓). In 2005, there were no significant differences in fruit composition at WV (data not shown).
Must and wine composition.
Soluble solids and titratable acidity were similar in the must from all treatments at SPV in 2004 (Table 5⇑). However, when berry skins, juice, must, and wine were compared at SPV in 2004, there were more total phenolics and antioxidant activity in skins and juice from berries in white vs. black geotextile plots, and the must from berries in control plots contained more total phenolics than that from berries in black geotextile plots (Table 6⇓). Differences in must composition among VFMS treatments were more closely correlated with anthocyanins and total phenolics values in finished wines than differences in berry skin or juice. The substantially greater phenolics content of juice compared with must samples was attributed to the fact that berries were frozen and heated to 80°C before extracting juice, while the must was obtained from fresh samples after crushing and destemming.
Components of yield and incidence of fruit rot.
At SPV, vines in white geotextile plots produced significantly more crop than those in control treatments in 2004, and a similar trend occurred in 2005 (Table 7⇓). Compared with the control treatment, vines in black geotextile plots had higher cluster weights in 2004, and those in white geotextile plots had higher cluster weights in 2005, but average berry weight was equivalent in all three treatments in both years. There were no differences in the incidence or severity of fruit rot among the VFMS treatments at either vineyard in either year (data not shown).
At WV, vines in white geotextile plots produced slightly greater yields both years, but these trends were not statistically significant (Table 7⇑). Cluster weights were greater in white geotextile plots during 2004 and 2005, but the only significant difference was that vines grown in white geotextile produced larger clusters than those grown in control plots in 2005. There were no statistical differences in the number of clusters per vine among VFMS treatments in either year at either vineyard.
Linear pairwise regression analysis for the components of yield in both vineyards for each year indicated that the yield per vine was closely correlated with the number of clusters per vine and the average cluster weight (Table 8⇓). Furthermore, cluster weight was more closely linked with the number of berries per cluster than with the average berry weight, suggesting that increased berry set may have been a factor in the yield differences among treatments.
Economic analysis.
Averaged over three years, the black and white geotextiles were substantially more expensive to install and maintain than the usual practice of mowed row middles and undervine herbicide strips used at these two vineyards (Table 9⇓). Costs of installing the row-to-row geotextiles at SPV were proportionally greater than the undervine geotextile strips at WV, and maintenance costs in the second and third years would also be greater at SPV because the row-to-row geotextiles were removed and replaced each year so soil could be hilled up around the vine trunks during winter. There was little difference in costs between the white and black geotextile treatments within each vineyard, and the costs of the control treatments were also similar at both sites. Despite the greater yields and crop values in white geotextile plots at both vineyards, the net gain from this treatment was less than that from the less costly standard control treatment (Table 9⇓).
Discussion
Light and temperature measurements.
The row-to-row white geotextile at SPV reflected substantially more sunlight up into the vines throughout the growing season than the black geotextile or control treatments (Figure 1⇑). The narrower white geotextile strip at WV reflected proportionally less sunlight, but still increased canopy light over the black geotextile or control treatments (Figure 2⇑). The white geotextile in our experiments reflected less light than an aluminized mulch used elsewhere in a Goblet-trained vineyard (Robin et al. 1996), but the light increases we observed over the white geotextile were similar to those reported elsewhere for mechanical and manual leaf removal (Percival et al. 1994a). As vine canopies increased in density and height each summer, less sunlight reached the ground or geotextile surface and reflectance was reduced. From veraison to harvest, the amount of reflected light was more consistent among treatments, with differences mainly due to variable weather conditions and cloud cover.
The negligible differences in average daily temperatures within vine canopies among the treatments in our two experiments were consistent with other observations that there were no differences in air temperature 5 cm above the surface of black or reflective mulches (Ham et al. 1993). Although black plastic mulches increased soil surface temperatures in their study, wind and convection dissipated the heat originating at ground level.
Vine size, winter bud survival, and nutrient analysis.
After two years of treatments, vine size at both vineyards was almost doubled in black or white geotextile plots compared to control treatments (Table 1⇑). We attributed the increased vine growth in both geotextile treatments to their suppression of weed competition throughout the growing season. Weeds recolonized the herbicide treated strip beneath vines in control treatment plots by late summer each year, and there was dense groundcover vegetation in the mowed row middles of control treatments. We did not measure soil water availability in these two nonirrigated vineyards. However, during the unusually wet and cool summer of 2004, it is unlikely that vine growth was limited by water deficits in any VFMS treatment. In contrast, during the hot, dry summer of 2005 it is very probable that weed competition for limited soil water reserves was greater in the control treatments, which could explain the increased vine size in geotextile plots that year. The reflective properties of white geotextile were presumably not linked with differences in vine size, because it was similar in both black and white geotextile plots at both vineyards.
Shading entire vines reportedly increased N and reduced Ca in leaves and increased NO3-N in petioles (Smart et al. 1988). Despite its opposite effects on canopy light levels, the white reflective geotextile had similar effects on petiole N and Ca in our study, and that was one of the few differences in vine nutrient status among the VFMS treatments in either vineyard (Table 2⇑). As with vine growth effects, we attributed the increased petiole N concentrations in 2005 to a reduction in weed competition with vines in both geotextile treatments. Despite the substantial differences in vine size and yields among the three VFMS treatments, there were surprisingly few differences in vine nutrient status.
Fruit composition.
Many studies have demonstrated that moderate weed competition for water and other essential nutrients during the latter part of the growing season can improve fruit quality in winegrapes (Ingels et al. 1998, Wheeler et al. 2005). In our experiments, the elimination of competing weeds and increased vine size in both geotextile treatments were probably responsible for the decreased soluble solids in fruit from these treatments (Table 3⇑). The lack of treatment effects on most other fruit attributes in our study was also consistent with recent studies of reflective mulches. When reflective mulch was applied for a shorter duration in late summer at an Ontario vineyard, there were no effects on soluble solids, pH, anthocyanins, or total phenolic concentrations of Cabernet franc, Cabernet Sauvignon, Pinot meunier, or Pinot noir berries (Pearson 2004). A similar lack of effects was reported on Merlot vines grown in two reflective mulches at a Massachusetts vineyard (Vanden Heuvel and Neto 2006).
Some previous studies found that various treatments increasing fruit exposure to sunlight subsequently increased soluble solids (Bergqvist et al. 2001, Coventry et al. 2005, Crippen and Morrison 1986, Kliewer et al. 1988, Rojas-Lara and Morrison 1989, Smart et al. 1988), while other researchers reported that increased light had no effects on fruit composition (Morrison and Noble 1990, Reynolds et al. 1995, Smith et al. 1988). Using leaf removal to increase light in the cluster zone was reported to reduce berry soluble solids because it left inadequate leaf area to ripen the remaining fruit (Zoecklein et al. 1992).
We observed relatively few treatment differences in berry pH or titratable acidity (TA) during this study (Table 3⇑), in accord with published findings (Pearson 2004). Another report noted that the effects of reflective mulch on pH and TA depended upon the variety of grapes involved (Robin et al. 1996). Others have reported that increasing sunlight exposure in the cluster zone can reduce juice pH and TA (Bergqvist et al. 2001, Kliewer et al. 1988, Rojas-Lara and Morrison 1989) or that it had little apparent effect (Crippen and Morrison 1986, Morrison and Noble 1990, Percival et al. 1994b, Smart et al. 1988, Smith et al. 1988). Reductions in TA were linked to lower malic acid concentrations in fruit exposed to direct sunlight or higher temperatures (Zoecklein et al. 1992). Considering the increased sunlight on clusters in reflective mulch treatments, it is surprising that many studies have shown no consistent effects on pH or TA in winegrapes.
Although the black geotextile treatments at SPV produced fruit with less anthocyanins, total phenolics, and antioxidant activity, the white geotextile had no effects on these fruit characteristics relative to control treatments (Table 3⇑). Reduced vine size has been associated with increased anthocyanins (Wheeler et al. 2005) and other phenolic compounds in grapes (Cortell et al. 2005), and the increased vine size in geotextile plots may have reduced anthocyanins and phenolics in grapes from black geotextile plots in our studies. In contrast, the increased sunlight on fruit in white geotextile plots may have enabled those vines to maintain anthocyanin and phenolics concentrations despite their increased vegetative growth. Cabernet franc and other varieties had higher phenolics concentrations when grown in aluminized mulch (Robin et al. 1996, Coventry et al. 2005). Other studies have also noted the importance of solar radiation on fruit for synthesis of anthocyanins and other phenolic compounds (Bergqvist et al. 2001, Crippen and Morrison 1986, Morrison and Noble 1990, Rojas-Lara and Morrison 1989, Smart et al. 1988, Smith et al. 1988).
Wine flavor evaluations.
Although one study found few other differences in standard measures of fruit composition, reflective mulch treatments in the vineyard were reported to reduce undesirable vegetable aromas in the resultant Cabernet franc, Cabernet Sauvignon, Pinot meunier, and Pinot noir wines (Pearson 2004). A taste panel of six experienced wine judges came to a similar conclusion when evaluating the three wines made from our treatments at SPV in 2004. Wine from the white geotextile plots lacked the herbaceous (bell pepper) defects noted in the other two wines. Vegetable aromas attributed to 2-methoxy-3-isobutylpyrazine are a problem in New York vineyards during years when weather patterns inhibit red varietal ripening in autumn, and this compound has been found in higher concentrations in shaded fruit of Sauvignon blanc and other winegrapes (Allen et al. 1996, Marais 1996, Smith et al. 1988). Unpublished studies involving reflective mulches for red winegrape varieties in New Zealand have shown similar improvements in wine aromatics (G. Creasy, personal communication) and this potential effect of reflective mulches deserves further study in cool-climate vineyards.
Components of yield.
These experiments showed a distinct trend of increased yields from vines in reflective geotextile plots (Table 7⇑). Regression analyses suggested that yield increases were primarily due to increased clusters per vine, but cluster weights were also strongly correlated with yield and differences in cluster weight were determined mostly by the number of berries per cluster (Table 8⇑). It appears that berry set increased for vines in the white geotextile plots, perhaps because of the increased sunlight on inflorescences during bloom and set (Figures 1⇑ and 2⇑). A reflective mulch reportedly increased yield and cluster weight in several varieties including Cabernet franc; the increased cluster weights were due to increases in both berry weight and berries per cluster (Robin et al. 1996). The light environment of shoots has been shown to affect bud fruitfulness and berry set in other varieties including Thompson Seedless (Sanchez and Dokoozlian 2005) and Sauvignon blanc (Kliewer et al. 1988); the potential yield increase from improved light levels during bloom and berry set in cool-climate vineyards merits further investigation.
We observed no differences in the incidence or severity of fruit rot (primarily caused by Botrytis cinerea) among the three VFMS treatments. In contrast, removing leaves to increase light in the cluster zone lowered the incidence of cluster rot in some other studies (Percival et al. 1993, 1994b, Smart et al. 1990, Zoecklein et al. 1992). Percival et al. (1993) also noted that fruit exposed to more sunlight had more epicuticular wax and thicker cuticle tissue layers in Riesling, Cabernet franc, and Optima grapes. It may be that leaf removal or shoot thinning are more effective for reducing cluster rot than merely increasing sunlight with reflective mulches—because leaf and shoot removal can also increase air circulation and permit more penetration of fungicides into the fruit zone of grapevines.
One factor that may have minimized potential vine responses to reflective mulch in our experiments was that vineyard managers at both sites thinned shoots and grape clusters during late July, removing most of the leaves surrounding fruit clusters (Figure 3⇓). The meticulous leaf and shoot removal in these two vineyards may have increased ambient sunlight in cluster zones enough to equilibrate the reflective mulch effects on canopy light environments.
Economic analysis.
Averaged over three years, both geotextiles were substantially more expensive to install and maintain than the standard practice of mowed sod alleys and herbicide strip in the vine rows (Table 9⇑). Because of the greater proportion of vineyard floor that was covered at SPV, the materials costs of geotextiles were substantially higher than at WV. Purchase costs of materials represented most of total VFMS expenses in the first year, while removing and reinstalling the geotextiles annually at SPV comprised most of their maintenance costs in subsequent years. Labor costs could be reduced by installing the geotextile in a 2‐m‐wide strip with a mechanical mulch applicator. The value of the increased yields from vines in white geotextile plots did not compensate for the much greater establishment costs for geotextiles compared with the standard treatment of herbicide strips under vines and mowed row middles. It would require substantial increases in wine quality and value to make the reflective geotextile an economically practical vineyard floor management system.
Conclusions
Reflective geotextile mulches may be an effective way to increase yields in coolclimate vineyards without reducing fruit quality. Geotextile mulches greatly increased vine size after two years, probably suppressing weed competition compared to the standard treatment. Had the vines in our two experiments been balancepruned in accordance with their higher pruning weights—leaving more fruiting nodes on vines in the geotextile plots—there could have been even greater differences in yield among the VFMS treatments. However, the reflective geotextile did not consistently advance veraison, increase soluble solids, increase pH, reduce titratable acidity, or increase anthocyanins and total phenolics in berries, must, or wine. Although the reflective mulch in our tests was more expensive than standard vineyard floor management practices, this additional expense could be acceptable if it improved the flavor chemistry and value of red grapes and wines from coolclimate viticultural regions.
Footnotes
Acknowledgments: The authors thank the USDA Viticulture Consortium–East and the N.Y. Wine Grape Foundation for funding this research.
We thank Dave Wiemann of Sheldrake Point Winery and John Wagner of Wagner Winery for letting us conduct experiments in their vineyards. Drs. Robert Pool, Leroy Creasy, Chris Watkins, Tim Martinson, and Andrew Reynolds provided helpful advice on the experimental methods, and Francoise Vermeylen was our statistical consultant.
- Received November 2006.
- Revision received May 2007.
- Copyright © 2007 by the American Society for Enology and Viticulture