Abstract
Cover crops can compete with vines for soil nutrients and thus can affect grapevine development and must and wine quality. The objective of this study was to evaluate the influence of two different cover crops on the availability of soil N, P, K, and Mg and on grapevine nutritional status, vigor, yield, and must and wine quality. The experiment was carried out in a cv. Tempranillo vineyard in La Rioja, Spain, using three treatments: a gramineous cover crop (barley), a leguminous cover crop (clover), and conventional tillage. Soil nitrate evolution and P, K, and Mg were determined, and total biomass and nutrient content of cover crops were measured. We also assessed leaf nutrient content, vine vigor, yield, and must and wine quality. Uptake of P, K, and Mg by cover crops did not reduce the soil availability of those nutrients and did not affect their concentrations in grapevines. The barley cover crop reduced soil N availability from the first year onward and led to decreased leaf N and vine vigor in the third year. Increased polyphenol content and color intensity were observed in the barley treatment in the fourth year, and these changes were more significant in must than in wine. The clover treatment increased soil N availability in years 2 through 4 and led to increased leaf N content in the third and fourth years. The use of barley as a cover crop could be a viable alternative for reducing soil N and improving must and wine quality; however, these effects required time to develop after introduction of the cover crop.
The AOC Rioja grapegrowing region, located in northeastern Spain, has a semiarid Mediterranean climate. Most vineyards in this area use tillage for soil-surface management to prevent weeds from competing for soil moisture. Soil tillage buries residues, disrupts macroaggregates, increases aeration, and stimulates microbial breakdown of soil organic matter (Reeves 1997). In part because of tillage, vineyard soils in the Rioja region have low organic matter content (<1%) and loamy texture (Peregrina et al. 2010) and thus are highly susceptible to erosion. Recent studies in Rioja indicate that the use of permanent cover crops can increase organic carbon, aggregation, and biological activity in vineyard soils, thus improving soil quality (Peregrina et al. 2010, 2014a, 2014b).
In addition to positive effects on soil quality, the establishment of cover crops in vineyards could be used to control vine vegetative growth and yield (Tesic et al. 2007, Ripoche et al. 2011). However, the presence of cover crops does not always lead to reduced vigor and yield or to enhanced must quality. Various studies have found minimal or no competitive effects of cover crops on grapevines and must quality (Ingels et al. 2005, Smith et al. 2008), and there are few reports on the effects of cover crops on wine quality.
Soil conditions affect cover crop growth and nutrient uptake. However, there is little information about competition between cover crops and grapevines for soil nutrients or about the effects of cover crops on soil nutrient availability and vine development.
Nitrogen (N) is important for vine vegetative growth and yield. Soil N availability in vineyards with no or low fertilization relies on adequate soil moisture and temperature to trigger the mineralization of organic N to nitrate-N (N-NO3−) (Pérez-Álvarez et al. 2013a), the main form of N taken up by vines. Previous studies have shown that cover crops can reduce soil N-NO3− availability in Mediterranean climates (Cellete et al. 2009, Ripoche et al. 2011, Peregrina et al. 2012). On the other hand, leguminous cover crops increase soil N availability and vine N uptake (Fourie et al. 2007b, Ovalle et al. 2010). Changes in soil N availability are important; in Rioja, a reduction of soil NO3− during bloom reduces yield, vine vigor, and must anthocyanin and polyphenol content (Pérez-Álvarez et al. 2013a, 2013b). A large percentage (~40%) of the N required for annual vegetative and reproductive vine development comes from the plant’s reserves (Williams 1991). Therefore, the impact of nutrient deficiency could appear several years after cover crops have been established.
There are few data about the dynamics of other nutrients, such as P, K, and Mg, in vineyard soils with cover crops. Grapevines require only small quantities of P (Conradie and Saayman 1989), but cover crops can compete for soil P and thus reduce vineyard P uptake. Potassium is an important macronutrient that regulates sap flow and grape acidity (Mpelasoka et al. 2003); cover crop competition for soil K could reduce K concentration and pH in grapes. This could be an advantage in vineyards that are rich in soil K and that thus have must with high pH and reduced grape color quality (Mpelasoka et al. 2003). High soil K can also result in poor flavor, sugar/acid balance, and wine stability; wines produced from crops grown with excess K require the addition of acid to prevent a flat taste (Rühl et al. 1992), and these musts and wines are highly susceptible to biological spoilage (Mpelasoka et al. 2003). Regarding Mg, the relationship between Mg status and sucrose export may explain observed negative correlations between petiole Mg content and berry sugar (van Leeuwen et al. 2004).
The objective of this study was to evaluate the effects of gramineous and leguminous cover crops on soil macronutrients (N, P, K, and Mg), vine nutritional status, vine vigor and yield, and must and wine quality in a high-vigor vineyard with cv. Tempranillo (the main variety in the AOC Rioja) during the first four years after cover crop introduction.
Materials and Methods
Experimental design and vineyard site description
The study site was located on a terrace of the Najerilla River in the La Rioja region (northern Spain; lat. 42°26′34″N; long. 2°43′32″W; 468 m asl). The vineyard was planted in 1999 to Vitis vinifera cv. Tempranillo on 110 Richter (110R) rootstock with 1.3 m × 2.7 m vine by row spacing. The soil was classified according to USDA Soil Taxonomy as Oxyaquic Xerorthent. Characteristics of the Ap horizon (0 to 15 cm) included 18.5% clay, 43.2% silt, 38.3% sand, 1.24% organic matter, and 4.2% carbonate content, pH 8.50 (1:5, 25°C), and electrical conductivity 0.11 dS/m (1:5, 25°C).
The vine rows ran east to west. Vines were trained on a vertical shoot-positioning trellis on double cordon Royat and were hand pruned to leave a two-bud spur (12 buds per vine). In spring, all water shoots were removed and left on the alleyway soil surface.
Weather conditions were recorded by a meteorological station of the Riojan Agroclimatic Service (www.larioja.org/siar) installed 2.28 km from the experimental vineyard (lat. 42°27′42″N; long. 2°42′45″W; 465 m asl). Precipitation during the growing season (April to October) was 166.8 mm in 2009, 209.4 mm in 2010, 176.8 mm in 2011, and 283.6 mm in 2012. Precipitation during the dormant season (November to March) was 251.2 mm (2009), 227.8 mm (2010), 180.2 mm (2011), and 137.0 mm (2012). Average annual temperature and potential evapotranspiration (FAO-Penman Monteith) during the growing season were 12.6°C and 808.7 mm (2009), 11.7°C and 760.4 mm (2010), 12.8°C and 804.6 mm (2011), and 12.5°C and 797.6 mm (2012).
The vineyard was not irrigated during the experiment. With the exception of soil management, standard cultural practices were the same for all treatments and were performed by the vineyard manager. The experimental vineyard had not been fertilized since 2000 (nine years prior to the experiment). Grapevine canopy management and control of vine pests and diseases by chemical spraying were carried out using machines, and berries were harvested manually. After pruning, canes were shredded and left on the soil surface between rows in all treatments.
Experimental design
The experiment used a randomized complete block design with three treatments and three replicates per treatment. Treatment replicates consisted of two adjacent alleyways on either side of a vine row, so each replicate (plot) consisted of four adjacent rows with 20 vines per row.
Three types of between-row soil cover management were studied: CT, conventional tillage; B, a barley (Hordeum vulgare L.) cv. Naturel cover crop; and CL, a Persian clover (Trifolium resupinatum L.) cover crop. At the start of the experiment (February 2009), all treatment rows were disked and rolled, and the cover crops were sown (50 kg seed/ha) into the prepared seedbeds. The cover crops were resown (50 kg seed/ha) in February 2011. The CL cover crop was mowed in spring (April) 2010, and the crop residues were left on the soil surface. Both cover crops wilted in summer (B in late June, CL in late July). To eliminate weeds in the CT treatment alleyway, soil tillage (0 to 15 cm) was performed every 4 to 6 weeks from February to August.
In all treatments, an 80-cm strip below the vines was kept free of weeds by applying herbicide consisting of glyphosate (36%, 0.75 L/ha) and pendimethalin (33%, 1 L/ha) with a backpack sprayer each year in the second week of February. In each plot, the alleyway between vine rows was 1.9 m wide, for the three treatments.
Soil sampling throughout the growing season
In each plot, three soil subsamples were randomly collected in the alleyway at 0 to 15 and 15 to 45 cm depths using an Edelman auger and were bulked into a composite sample. Five samples were collected during each growing season, at budbreak (late April), full bloom (100% of caps shed, mid-June), full fruit set (mid-July), full veraison (100% of berries colored, mid-August), and postharvest (early November). The soil samples were air dried, ground, and sieved to 2 mm. The percentage of coarse fragments (>2 mm) was determined in each sample.
Soil N-NO3− analysis
Soil N-NO3− was extracted from samples collected at each of the time points with 2 M KCl and was measured by colorimetry at 550 nm using a SEAL AutoAnalyzer 3 HR based on segmented flow (Seal Analytical, Hamburg, Germany). The N-NO3− content was expressed in kg/ha and was calculated from soil bulk density determined by the core method (Grossman and Reisch 2002) and the percentage of coarse fragments (>2 mm) at the 0 to 15 and 15 to 45 cm depth intervals.
Soil P, K, and Mg analysis
Phosphorous, potassium, and magnesium concentrations were determined in soil samples collected at bloom in each year (2009 to 2012). Assimilable P content was analyzed by extraction with 0.5 M NaHCO3 at pH 8.5 using the Olsen method (Olsen et al. 1954). Phosphorus in the resulting extracts was measured by colorimetry using a UV-visible spectrophotometer (Varian Cary 50 Conc; Varian, Inc., Cary, NC) and expressed as mg P/kg soil.
The soil samples taken at bloom were also extracted with ammonium acetate solution (1.0 N NH4C2H3O2, pH 7.0) (Thomas 1982). Potassium content in the extract was quantified by atomic emission spectroscopy, and Mg was quantified by atomic absorption spectroscopy (SpectrAA-140; Varian Ltd., Mulgrave, VIC, Australia). The K and Mg contents were expressed in cmol(+)/kg soil.
Cover crop biomass and nutrient uptake
Nutrient uptake by B and CL cover crops was estimated in the aerial organs at bloom in each year of the experiment. Aboveground plant biomass was collected in the vineyard alleyways in four randomly placed quadrats (0.25 × 0.25 m each) per plot. The plant material was cut at soil level and washed with tap water in the laboratory, rinsed with distilled water, and dried at 60°C in a forced-air oven for 72 hrs. The dry biomass was weighed and sieved to 0.5 mm in an ultracentrifugal mill (Retsch ZM1, Haan, Germany). Nitrogen concentration was determined by combustion in a CNS elemental analyzer (TruSpec CN, LECO, St. Joseph, MI).
Phosphorus, K, and Mg concentrations were determined by the microwave hydrogen peroxide digestion method and inductively coupled plasma (ICP)–optical emission spectroscopy (ICP-3300 DV; PerkinElmer, Waltham, MA). Total nutrient uptake was expressed in kg/ha.
Vine nutrient determination: leaf tissue analyses
Grapevine leaves were collected for nutrient analyses at bloom (mid-June) in each growing season. In each plot, 20 healthy leaves were randomly selected (one leaf with petiole per vine) and sampled from nodes opposite the first cluster (one leaf per shoot) (García-Escudero et al. 2013). The petioles were detached from the blades and both were washed with tap water, rinsed with distilled water, and dried at 60°C in a forced-air oven (Selecta DryBig; JP Selecta, Barcelona, Spain) for 72 hrs. The dried samples were ground and sieved to 0.5 mm in the ultracentrifuge mill (Retsch ZM1) and analyzed for macronutrients. Nitrogen concentration in leaf blades and petioles was determined with an elemental analyzer (TruSpec CN, LECO). Concentrations were expressed on a dry weight basis as g/100 g.
Phosphorus, K, and Mg contents were determined by microwave hydrogen peroxide digestion and ICP–optical emission spectroscopy (ICP-3300 DV, PerkinElmer).
Yield components, vigor indicators, and vegetative- productive vine balance
Each year, 18 vines were randomly selected and harvested in each plot just prior to the AOC Rioja commercial harvest. The number of bunches per vine and yield per plot (total harvest weight/ha) were recorded to calculate the average bunch weight (grapes/kg vine).
At postharvest (November and December), 10 vines were randomly chosen in each plot, and wood pruning weight and shoot number were determined. The Ravaz index was calculated by dividing the total yield per vine by the pruning weight recorded during the following winter.
Berry sampling and analysis
Each year, random samples of 500 berries were collected at harvest. Five clusters were collected from 20 grapevines distributed randomly throughout each experimental plot; five berries (two on opposite sides from the top of the cluster, two on opposite sides from the middle, and one at the tip of the cluster) were picked from each cluster. In the laboratory, 200 of these berries were separated, counted, and weighed to obtain the average berry weight. These berries were then crushed with the remaining berries using a masticator (IUL Instruments GmbH, Königswinter, Germany) to obtain must. Probable alcohol, pH, total acidity, malic acid, K, color intensity, and tonality of the must were determined according to the European Community Official Methods (Commission Regulation 1990); tartaric acid was determined by the Rebelein method (Lipka and Tanner 1974); and total polyphenol index (TPI) was determined by measuring the absorbance at 280 nm after conventional dilution of samples.
Winemaking
Each season, nine wines (three replicates per treatment) were produced in our winery. Approximately 80 kg of clusters harvested in each plot were destemmed, crushed, and placed into 100-L stainless steel fermentation vats. Alcoholic fermentation was carried out with active dry yeasts (Uvaferm VRB; Lallemand, St. Simon, France), at 19 to 21°C for 9 to 11 days, following traditional procedures used in AOC Rioja (including berry skins and seeds until the residual reducing sugar content was <2 g/L). After primary fermentation, malolactic fermentation was induced by inoculation with commercial lactic acid bacteria (Uvaferm Alpha “U,” Lallemand). Malolactic fermentation was completed when the concentration of malic acid was <0.2 g/L (after 8 to 10 days), and the wines were then bottled.
After approximately one month of cold stabilization (8°C) in bottles, wine samples were analyzed. Total acid, pH, tartaric acid, malic acid, K, and color parameters were measured according to the European Community Official Methods (Commission Regulation 1990).
Statistical analysis
Statistical analyses were performed with SPSS version 21 for Windows (SPSS, Inc., Chicago, IL). Normality and homogeneity of variance were evaluated with the Shapiro–Wilk and Levene tests, respectively. Treatment effects on the measured variables were tested using ANOVA (univariate linear model), and comparisons between treatment means were calculated using the least significant difference (LSD) multiple range test calculated at 95% (p < 0.05) and 90% (p < 0.1) confidence levels.
Results
Effects of cover crops on soil nutrient availability
Soil nitrate content, sampled each season at grapevine budbreak, bloom, fruit set, veraison, and postharvest, is shown in Figure 1. Soil N-NO3− availability was lower in B than in CT or CL at both soil depths throughout the four growth cycles. These differences were less pronounced at 15 to 45 cm and were significant with the following exceptions: at budbreak and bloom in 2010 and at bloom, fruit set, and veraison in 2012 (0 to 15 cm); and at budbreak, bloom, and fruit set in 2009, at bloom in 2010 and 2012, and at fruit set and veraison in 2012 (15 to 45 cm) (Figure 1A, B).
Soil N-NO3− availability (kg/ha) in each sampling year from budbreak (BB) in late April, bloom (BL) in mid-June, fruit set (S) in mid-July, veraison (V) in mid-August, and postharvest (P) in early November, in the 0 to 15 cm (A) and 15 to 45 cm (B) soil layers in the three treatments. CT, conventional tillage; B, cover crop of barley; and CL, cover crop of clover. Bars represent standard deviation. For a given phenological stage, different letters among treatments indicate significant differences (LSD test, p < 0.05).
In 2009, topsoil (0 to 15 cm) in CL had lower N-NO3− content than that in CT, but soil N-NO3− was higher in CL than in CT or B from the second year (2010) on (Figure 1A). A similar trend was observed for the 15 to 45 cm depth interval, in which CT had higher N-NO3− content than B or CL in the first year, while CL had the highest N-NO3− content from fruit set in 2010 to veraison in 2012 (Figure 1A, B).
Soil P, K, and Mg content at both depths at bloom are shown in Table 1. No significant differences in soil P were found between treatments at either depth. In 2009, exchangeable K in the 0 to 15 cm layer was significantly higher in CT than in B or CL. However, in 2012, CL had higher soil K at 0 to 15 cm than did B and CT. CL had significantly higher Mg content than the other treatments in the 0 to 15 cm layer in 2012, the fourth season after introduction of the cover crops.
Phosphorus, potassium, and magnesium contents in two soil depth layers in the three experimental treatments in four growing seasons.
Uptake of N, P, K, and Mg by cover crops
Dry biomass measured at bloom differed significantly between B and CL in 2009 (Table 2). From year two onward, N content at bloom differed between the cover crops, and in 2011, the N content in CL was four times higher than that in B. Nitrogen content in biomass was three times higher in CL than in B in 2010 and two times higher in 2012.
Nutrient contents at grapevine bloom in cover crop treatments and four growing seasons.
In 2009, uptake of P was significantly higher in B than in CL, but there were no differences in P uptake between treatments from 2010 onward (Table 2). Annual P uptake from 2009 to 2012 ranged between 1 and 4.7 kg/ha. There were no differences in K content of cover crop biomass between B and CL in the first season, but CL had significantly higher K content than did B in the second through the fourth year (Table 2). The B treatment had significantly higher Mg uptake than CL in 2009; this trend was reversed in the remaining sampling years (Table 2).
Effects of cover crops on grapevine nutritional status
In 2011, N content in both leaf tissues was significantly higher in CT and CL than in B (Table 3). In 2012, CL had higher blade N content than CT and B. In that same year, the highest value for petiole N occurred in CL, the lowest value in B, and intermediate values for CT. No differences in K, Mg, or P content of leaf tissues were observed between treatments, except in 2011 when P content was significantly lower in B than in CT and CL in both blade and petiole tissues (Table 3).
Nutrient content in leaf tissues at bloom in the three experimental treatments and four seasons.
Effects of cover crops on grape yield and vigor
Neither the number of bunches per vine nor grape yield (kg grape/vine) differed significantly between treatments (Table 4). Pruning wood and average shoot weight produced per vine were significantly lower in B than in CT and CL in 2011 and 2012 (Table 4). The Ravaz index ranged between 3.14 and 5.95 for the four years and was significantly higher in B than in CT and CL in all years except for 2010.
Grape and wood yield produced in the three treatments and four study years.
Influence of cover crops on must and wine properties
During the four years of analysis, there were no significant differences in probable alcohol, pH, total acidity, tartaric acid, or malic acid between treatments (Table 5). Must in B and CL had slightly lower K concentration than must in CT in all years. The B treatment had higher absorbance values (280 nm) than CT and CL in all years, but the differences were only significant in 2012 (Table 5). Similarly, color intensity was higher and must tonality was lower in B than in CT and CL in 2012.
Must analysis in the three treatments and four study years.
Most of the evaluated wine parameters did not differ significantly between treatments throughout the experiment (Table 6). In 2011, total acidity content in wine was higher in B than in CT and CL, and lactic acid content in wine was higher in CT than in cover crop treatments. In 2009 and 2011, tartaric acid content in wine was higher in B than in CT. Tonality was higher in B than in CT in 2009 and 2012 and CL in 2010. Similar to musts, TPI (absorbance at 280 nm) was significantly higher in B than in CT and CL in the fourth year.
Wine parameters in the three treatments and four study years.
Relationship between soil N-NO3− with N leaf and must polyphenols content
If the four years of the experiment are considered, negative significant correlations exist between soil N-NO3− at 0 to 15 cm and must polyphenols content (i.e., ABS 280 nm = 15.1 to 0.15 soil N-NO3−, r = −0.59, p = 0.0002, n = 36) and between N content in petioles at bloom and must polyphenol content (i.e., ABS 280 nm = 19.0 to 6.2% N, r = −0.56, p = 0.0004, n = 36).
Discussion
Effects of cover crops on soil N-NO3− availability
Soil N availability fluctuated throughout the growing season in all treatments, but the variation was smaller in the barley cover crop treatment, in which N increased from budbreak to veraison and then declined until postharvest.
Nitrogen availability decreased more under barley than under conventional tillage. Reductions in soil available N in vineyards with nonleguminous cover crops have been reported in California (Smith et al. 2008, Steenwerth and Belina 2008), in Mediterranean France vineyards (Celette et al. 2009, Ripoche et al. 2011), and in La Rioja, the region in which our study took place (Peregrina et al. 2012).
Cellete et al. (2009) suggested that reduced soil N availability could be due to lower soil moisture under cover crops, which would lead to a decline in N mineralization rates. Steenwerth and Belina (2008) also indicated that increased biological activity in the soil corresponded to increased microbial N-NO3− consumption. However, the lower soil N availability could be explained by competition for N between the grapevines and the cover crop because the lower soil N under barley at bloom than under conventional tillage was of the same order of magnitude as N uptake by barley into aboveground biomass.
The CL cover crop increased soil N more than did CT after the second year, which was explained by the ability of legumes to symbiotically fix atmospheric N2 (Ovalle et al. 2010). The amount of mineralizable N supplied by leguminous cover crops depends on the growth stage at which the cover crop is cut and incorporated into the soil (Kuo et al. 1996). The lack of increase in soil available N under the CL cover crop in the first year could be explained by the first mowing of CL having taken place in April 2010. Fourie et al. (2007a) reported higher soil inorganic N content under leguminous cover crops than under conventional tillage in surface soil (0 to 15 cm) in South African vineyards.
Effects of cover crops on P, K, and Mg availability
Cover crops had little influence on P, K, and Mg availability. Our results for soil P agree with those reported by Fourie et al. (2007a) for a South African vineyard in which availability of soil P was not affected by cover crops. In our study, soil P levels ranged between optimum and high according to guidelines presented by Sawyer et al. (2003). Therefore, soil P availability did not decrease the P nutritional status in any of the treatments.
Levels of soil K also ranged between optimum and high (Sawyer et al. 2003). The increase in soil K in the upper soil layer under clover could be due to residues of the cover crop left on the soil surface after mowing; clover residues were richer in K compared to barley (Table 2). Therefore, redistribution of soil K occurs due to uptake by cover crops and incorporation into the soil surface through residues of aerial biomass. This result agrees with Fourie et al. (2007a), who showed an increase in soil K under leguminous cover crops. The process that explains the increased K content in the topsoil could also explain the increase in soil Mg availability in surface soils after the fourth year under the CL cover crop.
Uptake of N, P, K, and Mg by cover crops
Cover crop N nutrient uptake depends upon whether the plant used as cover crop is leguminous or gramineous. Other researchers also found higher biomass in leguminous than in gramineous cover crops in vineyards in California (Ingels et al. 2005), Oregon (Sweet and Schreiner 2010), and Portugal (Lopes et al. 2011). Higher N content in leguminous cover crops than in resident vegetation was also reported in a California vineyard (Steenwerth et al. 2013).
The higher N concentration in CL than in B biomass would not be a result of an increased N soil extraction by the CL, but rather by its ability to fix atmospheric N. Ovalle et al. (2010) showed that a leguminous cover crop fixed atmospheric N, which contributed to grapevine nutrition in a cv. Cabernet Sauvignon vineyard in Chile. In the B treatment, all of the N in aerial biomass of the cover crop at bloom was extracted from available soil N and thus would not be available to the vines. Celette et al. (2009) also observed a decrease in soil inorganic N under a permanent cover crop. Regarding phosphorus, Ingels et al. (2005) similarly found no significant differences in biomass P content between treatments from the second year on. Clover took up more K and Mg than barley; legumes require greater quantities of K and Mg compared to graminoids (Marschner 1995). Ingels et al. (2005) also found higher K uptake in CL than in nonleguminous cover crops. Despite the higher K and Mg uptake by CL, there were no significant differences in the soil availability of these nutrients between CL and B.
Effects of cover crops on grapevine nutrient availability
Based on the reference levels described by García-Escudero et al. (2013) (Table 7) for cv. Tempranillo in the AOC Rioja at bloom, no nutrient deficits were observed. In the four years of observation, contents of N, P, and K in leaf blades were higher than optimal, and Mg levels were below the optimal range, except for year four.
Optimal nutrient levels for cv. Tempranillo in AOC Rioja (García-Escudero et al. 2013).
Results similar to ours regarding the effects of cover crops on N content in blades have been reported. Ingels et al. (2005) showed that leaf blade nitrate content for a cereal mixture with 50% barley (H. vulgare) and 50% oat (Avena sativa) were lower than those for a green manure mix with 35% bell bean (Vicia faba), 35% field pea (Pisum sativum), 20% common vetch (Vicia sativa), and 10% barley in a cv. Merlot vineyard in California. These researchers also found that leaf blade N content at bloom was lower under a cover crop of native perennial grasses than under a tillage treatment (Ingels et al. 2005). Sweet and Schreiner (2010) reported that leaf blade N content at bloom in a cv. Pinot noir vineyard in Oregon was higher under a clover cover crop than under tillage or a cover crop of annual winter species.
Nitrogen levels in petioles were within the optimum range for cv. Tempranillo under clover but were well below reference levels under B and CT (Tables 3 and 7). Phosphorus content in petioles at bloom were higher than optimal, Mg values were low, and K levels were within the optimum range for cv. Tempranillo in the fourth year under CT and CL (Tables 3 and 7). Similar findings for effects of CL and B cover crops on petiole N content have been described; Ingels et al. (2005) showed comparable results for blade and petiole N content (described above) in cv. Merlot in California. In an Australian cv. Chardonnay vineyard, Tesic et al. (2007) found that petiole N content decreased after three seasons with permanent resident vegetation compared to tillage. Steenwerth et al. (2013) reported that petiole N-NO3− was two-fold higher under tillage than under a cover crop of A. sativa in a mature cv. Merlot vineyard in Lodi, CA.
The reduction in leaf N content could be attributed to lower soil NO3− availability under barley that appeared three years after introduction of the cover crop. These results could be explained by the fact that in the first years under barley, the vines used N stored in their tissues (trunk and roots), but this N reserve was lost when leaves fell. Thus, after three years, N stored in the vines could not compensate for lower soil available N under barley, thus leading to differences in leaf-N status compared to conventional tillage.
Effects of cover crops on grape yield and vigor
Reduced vigor occurred in the B treatment compared to CT in the third and fourth years of the experiment (Table 4). Reductions in vigor in vineyards with cover crops (compared to tilled vineyards) have been found in various geographical areas, including in an Australian cv. Chardonnay vineyard with resident vegetation (Tesic et al. 2007) and in California cv. Merlot vineyards with cover crops of native grass (Ingels et al. 2005) and A. sativa (Steenwerth et al. 2013). However, cover crops did not reduce vine vigor in a cv. Chardonnay vineyard with cereal cover crops (Smith et al. 2008), in a cv. Pinot noir vineyard with native grasses (Sweet and Schreiner 2010), and in a cv. Cabernet Sauvignon vineyard with resident vegetation (Lopes et al. 2008). Ripoche et al. (2011) found a delayed effect of a tall fescue (Festuca arundinacea Shreb.) cover crop on vigor in a cv. Aranel vineyard, in which the reduction occurred in the second year. The longer time to vigor reduction in our study (in the third year) could be explained by the higher vigor of our vineyard: 1.58 kg/plant for our conventional tillage treatment versus 0.36 kg pruning mass/plant in Ripoche et al. (2011).
No reduction in yield was seen in the B treatment in the last two years of the experiment (Table 4). A lack of yield reduction combined with a decrease in vigor was also observed by Ingels et al. (2005) in a California cv. Merlot vineyard with a native grass and by Lopes et al. (2008) in a cv. Cabernet Sauvignon vineyard with resident vegetation in Portugal. In contrast, Tesic et al. (2007) (cv. Chardonnay vineyard with resident vegetation) and Ripoche et al. (2011) (cv. Aranel vineyard with F. arundinacea) found reductions in both vigor and yield. The stronger effect of cover crop on yield observed by Ripoche et al. (2011) could be due to the lower vigor in this vineyard compared to ours.
The Ravaz index increased in the B treatment as a result of the decreased pruning weight. Because of these changes, from year three on, the Ravaz index for barley was within the optimum range of 4 to 7 proposed by García-Escudero et al. (2006) for balanced vines of cv. Tempranillo in AOC Rioja. However, the CT and CL treatments had Ravaz index values <4, which indicated unbalanced vines with excessive vigor.
Influence of cover crops on must and wine properties
Titratable acidity showed a decreasing (but nonsignificant) trend in must in the barley treatment compared to must in the conventional tillage and clover treatments. In agreement with our results, Ingels et al. (2005) and Sweet and Schreiner (2010) found no significant effects of cereal cover crops on titratable acidity in Merlot and Pinot noir cultivars, respectively. Lopes et al. (2008) reported lower titratable acidity under a cover crop of resident vegetation in cv. Cabernet Sauvignon, and Smith et al. (2008) found a decrease in titratable acidity in cv. Chardonnay with a cover crop of cereals. Lopes et al. (2008) explained their findings by the indirect effects of mild water stress on vegetative growth that could have improved the bunch microclimate. Under our experimental conditions, the reduced vegetative growth under barley in the third and fourth growing seasons could have been insufficient to improve the bunch microclimate and therefore to reduce titratable acidity.
The high levels of soil available K could explain why uptake of K by the cover crops did not affect the K status of vines or must. Higher levels of must polyphenols and color intensity in the fourth year under barley compared to conventional tillage and clover coincided with lower vine N content under barley. Over the four years of the experiment, significant negative correlations were found between soil N-NO3− (0 to 15 cm) and must polyphenol content. These results agree with relationships between N and must polyphenol content observed in other studies. Delgado et al. (2004) found that application of N fertilizer decreased the accumulation of phenolic compounds in berry skins of cv. Tempranillo. In addition, among various vineyard soils, low vine N status induced high polyphenol content in cv. Cabernet Sauvignon (Choné et al. 2001) in the Bordeaux region (France). Pérez-Álvarez et al. (2013a) found a negative correlation between soil N-NO3− concentration at bloom and polyphenol and anthocyanin content in cv. Tempranillo berries in a tilled vineyard adjacent to our experimental vineyard.
The wines obtained from each treatment were within the standards for wine parameters in AOC Rioja. In agreement with the minor effects of cover crops on must composition, the wines did not differ significantly (p > 0.05) between treatments in any of the analyzed parameters. Our results agree with those reported by Fourie et al. (2007b) for a South African Sauvignon blanc/Ramsey vineyard with cover crops of grains or N-fixing broadleaf species. However, at a lower significance level (90%, p > 0.1), wine and must polyphenol contents were higher in the fourth year with the barley cover crop compared to conventional tillage or clover. As discussed above, the increased polyphenol content could be due to lower vine N status, a consequence of lower soil N-NO3− availability throughout the four growing seasons. Increased polyphenol content under the barley cover crop treatment would have a positive effect on wine quality through improvements in color, organoleptic characteristics, and stability.
As with must, wine from the clover cover crop treatment did not differ from that of conventional tillage, which indicates that clover does not improve wine quality under the conditions studied here.
Conclusions
Cover crops had different effects on nitrogen availability compared to conventional tillage. The use of barley led to reduced N availability from the first year, while the use of clover increased soil N after two years. Clover increased vine N status but did not affect the properties of must or wine. Changes in vine N status under barley were reflected in the must four years after the cover crop was established and led to increased polyphenol content and color intensity. Although these changes in must affected the wine, the effects were not significant relative to the control (conventional tillage). The use of gramineous cover crops such as barley under the soil and climatic conditions of the Rioja region could be a useful approach for reducing N availability and, thus, for reducing vigor and increasing the polyphenolic content of grapes. On the other hand, clover cover crops could be used to improve N levels in vineyards with soil N deficits.
Acknowledgments
This study was supported by the Spanish Ministry of Science, INIA (Instituto Nacional de Investigación y Tecnología Agraria y Alimentaria), and European Social Fund through project INIA-RTA 2009-00101-00-00. F. Peregrina thanks the INIA and European Social Fund for his postdoctoral grant and E. Pérez-Álvarez thanks the INIA for her predoctoral grant. The authors particularly thank Ricardo Leza for lending them the vineyard in which they installed the experimental plots and for helping with the field labor.
- Received July 2014.
- Revision received December 2014.
- Revision received February 2015.
- Accepted February 2015.
- ©2015 by the American Society for Enology and Viticulture