Regulated deficit irrigation as a water management strategy in Vitis vinifera production

R.L. Wample, Dept. of Viticulture and Enology,
California State University, Fresno, California, United States of America
R. Smithyman, Stimson Lane Vineyards and Estates, Paterson, Washington, United States of America

Summary

An initial six-year study in a commercial vineyard located in the Columbia River Valley of Washington State, United States of America, examined the management practices and potential benefits of regulated deficit irrigation (RDI) on Vitis vinifera cv. Sauvignon blanc. The objective of the treatments was to evaluate the effect of deficit irrigation prior to, compared with after, veraison. Each of four irrigation treatments was applied to 1.6 ha and replicated four times for a total 27.0 ha. Irrigation treatments were based on desired soil moisture levels in the top metre of the profile where most of the root system is found. Soil moisture was monitored using a neutron probe and the information was combined with calculations of evaporative demand to determine the irrigation required on a weekly basis. Vine growth, yield, fruit quality and cold hardiness were monitored throughout the study. The results indicated that RDI prior to veraison was effective in controlling shoot growth, as determined by shoot length and elongation rate, as well as pruning weights. Sixteen wine lots, each of approximately 12 000 litres, were prepared each season. Although there was some effect on berry weight, yield was not always significantly reduced. Full irrigation prior to veraison resulted in excessive shoot growth. RDI applied after veraison to vines with large canopies resulted in greater water deficit stress. Fruit quality was increased by pre-veraison RDI compared to post-veraison RDI based on wines made. Regulated deficit irrigation applied at anytime resulted in better early-season lignification of canes and cold hardening of buds. There was a slight improvement in mid-winter cold hardiness of vines subjected to RDI. However, this effect was inconsistent. Studies on Cabernet Sauvignon and White Riesling are underway to confirm these results and to investigate the impact of RDI on fruit quality and winemaking practices.

The introduction of grapevines, especially Vitis vinifera cultivars, into new growing regions, has led to an increasing focus on irrigation to maintain or increase vine productivity and fruit quality. Irrigation and the developing strategies for using irrigation as a management tool in winegrape production have been ongoing for at least 20 years. Therefore, an understanding of plant water relations and soil water management is essential to use irrigation successfully to produce consistent yields of high-quality grapes. Effective irrigation management results in better control of plant growth and more efficient and economical crop production. A number of studies examining the effects of decreasing levels of irrigation on vine growth and physiology (Smart and Coombe, 1983; Mullins, et al., 1992; Williams and Matthews, 1990; Goldberg et al., 1971) have found that drip irrigation resulted in improved WUE based on fruit production and pruning weights per unit of water applied.

In 1972, Peacock et al. (1977) compared drip, sprinkler and flood irrigation, and found that drip used less water while achieving good vine vigour, fruit production and quality. However, there was evidence of salt accumulation in a smaller wetted rootzone compared to flood and sprinkler irrigation. In 1985, Bucks et al. (1985) reported similar results for the production of table grapes in Arizona. Finally, Araujo et al. (1995a, 1995b) found that similar crop production of Thompson Seedless could be achieved with either furrow or drip irrigation. However, they reported a reduction in the nitrogen content of drip irrigated fruit and they found a restricted rootzone associated with daily applications of drip irrigation. This led them to propose the use of drip irrigation to control vine vigour by restricting nitrogen uptake and restricting root volume.

In a study examining the potential benefit of using drip irrigation on Concord vines trained to various trellis systems, Cline et al. (1985) found that during dry years in New York, United States of America, drip irrigation improved yields, especially on higher density plantings and with trellis-training systems with higher cropping potential. They found drip to be more compatible than sprinkler irrigation on heavier clay soils with lower infiltration rates. Neja et al. (1977) found that timing of irrigation when combined with variations in trellis type resulted in differences in yield and quality of Cabernet Sauvignon grown in the Salinas Valley, California, United States of America. However, their results showed a higher yield with an intermediate level of irrigation when combined with a more elaborate trellis system.

Bravdo and Hepner (1987) showed that drip irrigation was an effective way to apply fertilizer to grapevines with the potential for influencing fruit and must composition. They found a significant response to phosphorus applied through the irrigation system, with higher yields, higher cluster numbers, improved wine sensory characteristics, wine colour and monoterpene levels in the must. They suggested that restriction of the rootzone by irrigation management could be used to control grapevine vigour. Irrigation systems have also had limited success for the application of herbicides (Fourie, 1988).

Irrigation management as a tool for use in the production of grapes has continued to receive attention in many regions of the world. In Australia, the use of regulated deficit irrigation (RDI) has been explored to control vegetative growth and improve the consistency of fruit production and quality (Goodwin and Jerie, 1992). In Spain, Nadal and Arola (1995) reported increased yield, malic and total acidity, and earlier ripening of irrigated Cabernet Sauvignon.

There is a growing need and desire to understand the effects of irrigation (water management) on grapevine growth, development, productivity, and fruit quality. The continued expansion of agriculture, including grape production, into low-rainfall regions compels a response to these issues. Furthermore, increased competition for this increasingly scarce resource will impose greater efficiency in irrigation management practices.

When considering using irrigation management as a tool, it is essential to establish a clear set of goals and to determine where water management may have an impact on them. Possible goals may include controlling vine vigour, preventing occasional water deficit stress, attempting to manage fruit development (berry size), or attempting to alter fruit quality by influencing soluble solids, pH, or titratable acidity. Careful selection of the most appropriate irrigation system for the vineyard site is also a high priority. The irrigation system must match soil type, depth, water holding capacity, infiltration rate, and the effective rooting zone of the vines. This latter point may require detailed knowledge of the cultivar or the rootstock in question. The amount of water available and its cost also demand careful consideration. Vineyards planted on hillsides or rolling terrain are not amenable to furrow or flood irrigation practices. Soils with low infiltration rates and significant slope also present runoff problems for overhead sprinkler systems with high delivery rates. Drip irrigation can accommodate all of these situations, but has higher initial capital investment costs and generally requires a higher level of management. Additional factors that warrant consideration are water quality, filtration requirements, system automation, and local availability of equipment, supplies, and support. Because of the number of variables involved, growers should contact companies dealing in irrigation design and equipment for recommendations tailored to the vineyard site.

Deciding when and how much to irrigate requires a thorough understanding of the factors that contribute to vine water status and the effects of various water management strategies on grapevine development and productivity. Grapevine water stress develops when the supply of water from the soil through the root system to the growing shoots is less than the evaporative demand. The cause for this imbalance may be: low available soil moisture; a poorly developed, injured, or otherwise restricted root system; unbalanced development of shoot and root systems; and/or high evaporative demand conditions. Salts in the irrigation water or in the soil can also reduce the water available to vines. Extensive trellis systems may contribute to leaf exposure and consequently a higher rate of transpiration than can be supplied by the roots. This latter type of stress is more likely to be transient in nature and less of a concern when adequate soil moisture is available.

The following are some observations and comments regarding the response of grapevines to water stress/management. Some of these statements are supported by research, whereas others are observations that appear to be consistent over several production regions.

Grapevines, especially Vitis vinifera, do not generally exhibit immediate signs of water stress, but will show symptoms of repeated stress by cumulative effects on shoot or fruit development. Williams et al. (1994) reviewed the effects of water stress and other environmental factors on grapevines. Depending upon the phenological stage at which it occurs, water stress has a wide range of effects on grapevine growth, development and physiology.

Water stress occurs infrequently during bud break and early shoot development due to low water use. However, water stress during this time may result in uneven bud break and stunted shoot growth. More severe and prolonged water stress may result in poor flower-cluster development and reduced pistil and pollen viability and subsequent berry set (Hardie and Considine, 1976). Nutritional deficiencies, especially in N, Mg and Ca, might also become evident under severe water stress (Falcetti et al., 1995). Most of the nutrients required by grapevines early in the season are derived from stored sources, thus reducing the likelihood of early-season deficiency symptoms. Early-season deficiencies in Zn and or B are often the result of water stress the previous season, causing reduced root growth and nutrient uptake (Christensen, 1962).

Following berry set, severe water stress may cause flower abortion and cluster abscission, possibly associated with hormone changes (During, 1986). Uncorrected water stress during this stage of development may result in reduced canopy development and, consequently, insufficient leaf area to adequately support fruit development and maturation. Because initiation of clusters at nodes 1-4 for the following season begins about two weeks prior to full bloom and continues for about two weeks, water stress during this stage may reduce the following season's crop potential. The predominant effect at that stage is believed to be a reduction in the number of clusters per shoot and not the number of flowers per cluster, which develop later in the season and throughout the dormant season as conditions permit.

Immediately after fruit set, water stress may restrict berry cell division and enlargement, resulting in smaller fruit and lower yield. The lag phase of berry development, which follows early berry development, is less susceptible to water stress. However, shoot development, which normally continues during this stage of development, would be reduced by water stress. Insufficient canopy development during this time will limit the photosynthetic capacity of the vine and may restrict fruit development and quality. Aside from reduced yield potential and fruit soluble-solids accumulation, the fruit may have higher pH, decreased acidity, and reduced colour development in red varieties. Problems associated with fruit sunburn are also more likely.

Rapid senescence of lower leaves, leaf abscission, and progressive loss of canopy, are consequences of water stress that may occur at any stage of development, but are more likely where a larger canopy is present. Sunburn of both red and white varieties can be a consequence of sudden fruit exposure caused by senescence of lower leaves and sudden loss of canopy and reduced canopy cooling caused by low evapotranspiration. Slow development of stress is associated with a loss of acidity and a rise in pH and soluble solids. More rapid onset of stress causes these processes to be arrested as fruit dehydration and raisining occur. Late-season water stress contributes to acclimation of one-year-old wood that begins from the base towards the tip of the cane. High levels of stress will result in abscission of shoot tips, which, if followed by over-irrigation, may stimulate lateral shoot growth. Such growth creates a competitive sink for photosynthates and delays fruit maturation. Late-season irrigation, following water stress, can also reduce cane and vine acclimation increasing the potential for low-temperature injury. Such vines are unlikely to have adequate viable buds the following season. Were exposed to extremely low temperatures, they often show reduced survival of buds, trunks and cordons.

The most detrimental effect of water stress following harvest is the potential for reduced root growth, resulting in decreased nutrient uptake and micronutrient deficiencies the following spring. Low-temperature injury of roots is also a concern if the soil remains dry, thereby increasing the depth of frost during long periods of cold weather. This is more likely in areas with lighter soils and little or no precipitation prior to winter conditions. Root injury is often expressed the following spring as delayed and erratic bud break, and eventual collapse of the developing shoots.

Careful water management is recognized as a tool for achieving some control of grapevine growth and development. The adoption of such a management strategy involves moderate stress at specific stages of development to achieve specific results. The decision to use such an irrigation strategy requires well-defined goals including effects on yield, grape quality, canopy structure, and protection against winter injury. To achieve these goals in the face of variable weather conditions requires both a thorough understanding of the effects of water stress on grapevines at various phenological stages, and also a good understanding of soils and soil water management. This understanding must include knowledge of total and available water holding capacities of the soil and the potential, as well as actual, rooting depths. The role and water use characteristics of cover crops also require careful consideration. Where most of the available moisture during root development is from irrigation water, irrigation methodology and scheduling can influence the distribution of the roots, both vertically and horizontally. Vineyard managers should be familiar with the characteristics of the rootstocks they use.

It is possible to use established crop coefficients (Kc) and measurements or calculations of potential evapotranspiration to estimate water use by vines (ET_p). Grapevine crop coefficients have been developed in several different locations (Evans et al., 1993; Grimes and Williams, 1990; FAO, 1977) and reflect the development of leaf surface area and vine water demand as the growing season progresses. The Kc represents the fraction of the potential evapotranspiration used by the vines, and its value is typically less than one. Variability in vine development from year to year has resulted in referencing the values of Kc with accumulated growing-degree-days (GDD) rather than calendar dates. Crop coefficients are low early in the season due to small leaf area and hence low water use, and approach unity as the canopy reaches maximum development in July and August in northern climates (January and February in the southern hemisphere). The calculation of daily water use (DWU) uses the published Kc for the appropriate accumulated GDD multiplied by the ET_p, a value based on the water use of a well watered, mowed, grass-covered area:

DWU = ET_p x Kc

The availability and use of computers make these calculations and record keeping easy, and facilitate improved water management. The calculations must also account for any rainfall that occurs during the irrigation cycle. It is important to recognize that not all rainfall reaches the vine's rootzone, and may, therefore, be considered as effective rainfall.

On a worldwide basis, the estimated range for total water use for wine-, table- and raisin-grape production, with or without irrigation, might vary from 10 to 31 ha-cm/year. Recognizing that grapevine water use increases through the season to a peak shortly after veraison, it is possible to further estimate the fractional water use during the major phenological stages (Table 1).

Table 1
Water use by stage of development

Using these estimates in conjunction with the annual precipitation for a given geographical location provides a sound basis for determining when and how much irrigation may be necessary in that area. It is also necessary to consider the suitability of the precipitation pattern for grape production.

Materials and methods

The study took place in the period 1992 - 1997, with the objective of evaluating the potential of using irrigation management to control wine grape vegetative growth and development, while maintaining yield and potentially improving fruit quality. The general approach to achieving these goals was to utilize the inherent growth characteristics and physiology of Vitis vinifera in combination with various irrigation schedules.

The vineyard site is in the rain shadow of mountains, and consequently receives about 20-25 cm of precipitation per year; the majority occurs between October and April. Rainfall during the growing season is considered to be ineffective as it occurs in small amounts and is frequently followed by high winds that increase evaporative demand. The number GDD (base 10°C) accumulated at the vineyard site averaged 1 600, with extremes of around 1 400 and 1 800.

The vineyard, 27 ha of Sauvignon blanc, is located at the Columbia Crest Estate Vineyard and Winery in south-central Washington, near Paterson, United States of America,. It was planted in 1979 with a 3.1x1.8 m spacing. Vines were trained to a bilateral cordon system, and spur pruned. The vineyard, originally irrigated using a centre-pivot system, was converted to drip irrigation in the mid-1980s. Drip irrigation lines consisted of pressure compensating 2-litre/h emitters at a 100-cm spacing.

Irrigation strategies were applied from 1992. They involved high irrigation (H) defined as 5.6 cm of water and low irrigation (L) defined as 3.1 cm of water per 30 cm of soil in the top 1 m of the soil profile. The H treatment is near field capacity, while the L treatment is near the permanent wilting point for the Quincy soil type found on most of this vineyard. The strategies were:

• (HL) High irrigation applied early in the growing season followed by low irrigation from the point when control of canopy development was achieved in the early-season low-irrigation treatment. The high-irrigation treatment phase typically occurred from bud break to early or mid-July, which is similar to the standard irrigation practice for wine grapes in the State.
• (HH) High irrigation maintained throughout the growing season. This extreme treatment was applied primarily for comparative purposes. However, it was suspected that some growers were using this practice.
• (LL) Low irrigation applied throughout the growing season. This extreme treatment was applied primarily for comparative purposes.
• (LH) Low irrigation applied early in the growing season until control of canopy development was achieved, which typically occurred by early to mid-July, followed by high irrigation through harvest.

At the end of each season, all treatments were irrigated to bring the top 40-60 cm of soil to near field capacity. This provided winter protection for the root system and adequate moisture for early season growth the following year. It was anticipated that there would be sufficient precipitation during the dormant season to fill the soil profile to the 1-m level. Where adequate precipitations did not occur, additional irrigation was applied prior to, or during, budbreak, to fill the soil profile to a depth of one metre.

There were four replicates of each treatment of 1.6 ha each. Treat-ments were randomized within each 6.5-ha set of replicates (Figure 1). Each 1.6-ha replicate was irrigated independently and equipped with a flow meter. Sixteen neutron-probe sites within each replicate (128 in total) provided weekly readings during the growing season. At each of the neutron-probe sites, four vines were selected for collection of data on growth and yield, i.e. a total of 512 data vines. Vines were spur pruned, leaving 36-40 buds per vine for all treatments. Pruning weights were taken from designated plot vines and randomly selected vines within each treatment-replicate.

Figure 1
Experimental plot design of the Clore vineyard irrigation experiment

Results and discussion

Over the six years of the study, soil moisture measurements taken during the third week of April indicated no differences among treatments in the top 1 m of the soil. Hence, there was uniform soil moisture for early-season growth, and a uniform starting point for irrigation planning. Irrigation schedules were determined using a combination of established Kc for winegrapes in Washington,United States of America (Figure 2) and (ET_p). The product of these values estimated the actual evapotranspi-ration ET_a. The ET_a and measured soil moisture were used to determine the hours of irrigation to achieve the desired soil moisture. The average cumulative irrigation over six years for the HH vines was about 50 ha-cm. The average for the LH vines was about 40 ha-cm, while the LL vines received about 30 ha-cm and the HL vines about 36 ha-cm. The HL vines received only about 2.5-5.1 ha-cm more than the LL vines because of low irrigation require-ments early in the season (the Kc and ET_p values are lower during April - June than in July and August). Although the HH vines received 50 ha-cm of water, this was still con-siderably less than the 76-90 ha-cm typically used previously in the Yakima Valley. The data provided in Figure 3 are indicative of the amounts of water applied annually to each of these treatments.

Figure 2
Crop coefficient for determining irrigation requirements

Modified from Evans et al. 1983

Figure 3
Cumulative irrigation applied to the four treatments in 1995

Irrigation maintained the HH and HL soils near 5.6 cm of water per 33 cm of soil until the first week of July, whereas the moisture in the LL and LH soils declined to 3.0 cm of water per 33 cm of soil in the top 1 m of the soil profile as determined by neutron-probe measurement. Irrigation, based on weekly consumption, was applied to maintain this level of soil moisture. Once shoot growth decreased in the LL and LH vines, the transitions in irrigation treatments were made, generally during the first or second week of July (Figure 4).

Figure 4
Soil moisture profiles determined by neutron-probe analysis 1992-94

Measurements of shoot length and node number were taken from data vines located near the neutron-probe sites. Individual shoots were selected on each cordon and marked for repeated measurements taken on a weekly basis. Shoot length of current-season growth was measured from the junction with 1-year-old wood to the shoot tip. Node number included all nodes from the base of the shoot to the last discernable node at the tip of the cane.

Leaf area measurements, achieved non-destructively by measuring the widest part of the leaf, were regressed against measured leaf area each season as determined with a LiCor leaf area meter. Repeated measurements taken weekly from the same shoots and leaves provided an indication of dynamic vine growth and development.

Fruit and wine quality analyses were based on harvesting the fruit from the various irrigation treatments at 23 percent soluble solids. All replicates were sampled and data were kept separate for statistical purposes. Harvest was based on the average of all replicates for a treatment reaching 23 percent soluble solids. Harvests commenced at about 2100 hours, with completion by 0900 hours the next day. The fruit of each replicate was kept separate for yield and winemaking purposes, thus allowing statistical analysis of these large-scale plots. Fruit from each treatment-replicate was crushed and pressed separately and a 114-h/litre sample placed in separate fermentation tanks. Plot vines for any given treatment were hand-harvested before mechanical harvest. Cluster counts and weights were based on hand-harvested fruit. Post-harvest soluble solids measurements were based on samples taken from the fermentation tanks for each of the four replicates of each treatment.

Measurements of shoot length, node number and pruning weight all demonstrated the ability to control shoot growth by irrigation management. Plates 1 and 2 represent typical differences in canopy development between the HH and LL irrigation treatments.

Plate 1
Canopy characteristics of the high irrigation (HH) treatment, August 1996

Plate 2
Canopy characteristics of the continuously low (LL) treatment, August 1996

Data indicate that, regardless of the previous year's irrigation, there was essentially no difference in shoot length from bud break until approximately 30 d after bloom. This was despite significant differences in weather conditions over the four years and the lower irrigation in the LL and LH treatments. This suggests several things. First, water was not a limiting factor early in the season. Second, because there were nearly 20 nodes present by the time differences in shoot growth developed, there was sufficient leaf area to mature the crop. Third, as even the HH irrigation treatment showed a change in shoot length around 30 d after bloom, fruit set and early cluster development reduced shoot growth. In general, the HL vines stopped initiating nodes shortly after the change in irrigation treatment, whereas there was an increase in shoot growth in the LH vines. Leaf area measurements also showed that the sizes of leaves up to about leaf-number 15 were similar regardless of irrigation treatment. This further supports the suggestion that early-season soil moisture was not limiting and that there was little difference in the water status of the irrigation treatments until late June or early July. Leaf area development was more sensitive to soil moisture depletion with differences occurring 10-14 d prior to differences in shoot elongation. Thus, changes in leaf area enlargement can serve as an early indicator of soil moisture depletion and, if carefully monitored, may be of use in scheduling irrigation. By mid- to late August, there was interior leaf senescence and defoliation in the HL treatment associated with water stress. However, in the LL vines there was less leaf senescence, indicating physiological adjustments resulting in increased WUE. From late July until the end of the season, following the change in irrigation, although there was no difference in soil moisture between the LH and HH treatments, the LH vines showed less stress as indicated by leaf and xylem water potential measurements. This is presumed to be the result of smaller canopy and physiological adjustments associated with the early-season low irrigation. Although the LL treatment often showed high levels of water stress, these vines showed less leaf senescence and loss than the HL vines.

The irrigation treatments had little effect on the number of clusters. The similarity in cluster number indicates that early-season low irrigation was not detrimental to the cluster initiation process.

During the first year of the study, which was very hot, the LL and LH vines averaged 1.5-2 t/ha less yield than the HH vines (Table 2).

Table 2
Fruit yield for each of the four irrigation treatments

Higher yields in the HH vines were due primarily to larger berries. Throughout the study, there tended to be more berries per cluster in the LL and LH treatments.

Fruit and wine-quality analyses were based on harvesting the fruit from the different irrigation treatments at the same soluble solids content. Post-harvest soluble solids measurements were based on samples taken from 114-hl tanks for each of the four replicates of each treatment. There were no differences among treatments in any of the five years of the study (Table 3). In 1995, the HH treatment was harvested at about 22 percent soluble solids, while the other treatments were all near 23 percent. Late-season high irrigation (HH and LH) tended to delay harvest and lower the soluble solids slightly, throughout the study. In 1993, the HH vines were harvested nearly a week after the LH vines which tended to be the first to reach 23 percent soluble solids. In cool, wet years like 1995, earlier harvest can be an advantage by avoiding fruit-rot problems.

Table 3
Influence of the four irrigation treatments on fruit soluble solids

The eitratable acidity of 1995 tank samples was significantly higher in the HH and LH must (1.0) than in the HL and LL must (0.7) (Table 4). Although this was due in part to the lower soluble solids for the HH treatment in 1995, this trend was seen in at least four of the five years. The lack of significant effects in 1992 was probably due to the high temperatures that prevailed throughout the season. Differences were generally accompanied by lower pH in the HH and LH musts than in the HL and LL musts (Table 5). Fruit and must analyses over the past five years have shown similar results. The lack of differences in soluble solids, while consistent differences occurred in pH and acidity, seems to indicate the effect of irrigation practices on these fruit, and potentially on wine characteristics.

Table 4
Influence of the four irrigation treatments on fruit titratable acidity

Table 5
Influence of the four irrigation treatments on fruit acidity (pH)

Vine evaluation during early August typically showed that treatments involving reduced irrigation had more lignified nodes than did HH. This was consistent over the five years of the study, indicating better cold hardiness during late summer and early fall. Although not important in most years, it could be a significant advantage in a year with an exceptionally early killing frost. Evaluations of cold hardiness of buds, undertaken each year from October to March, indicated no differences as a function of irrigation treatment.

The information produced by this study demonstrates that, given the variety and location, it is possible to produce a satisfactory crop of winegrapes with between 30 and 50 ha-cm of water per year including a post-harvest irrigation to bring the soil to a moisture level that will protect the root system from cold injury.

Several points from this study are applicable to vineyard water management in general. First, the water requirements of grapevines change as the season progresses and, second, their responses to changes in water availability at different stages of development are an important consideration.

The decision to adopt the concept of irrigation as a management practice should be based on well-defined objectives and on a clear idea of how irrigation management will overcome any problems. Where the problem is vine water stress, and irrigation water is available, the question is one of economics associated with the installation of an appropriate irrigation system and the expected improvement in vine growth and productivity. Depending upon when and why the stress occurs, soil and site characteristics, and grape variety, the decision to irrigate and the choice of irrigation system will vary significantly.

Based on information derived from this study, the only vine-related expense of using regulated deficit irrigation as a management tool is a potential loss of yield if stress becomes excessive. Leaving more buds at pruning can compensate for this, although it would be preferable to improve water management. Other costs are those associated with establishing, maintaining and operating the irrigation system. These costs require careful evaluation based on the potential for more consistent, balanced vine growth and fruit production of higher quality that would result in higher net returns to the grower. In addition to direct improvements in fruit quality, additional benefits observed in this and other studies include improved control of disease and pests. This is associated with a more open canopy that is less susceptible to pathogens and insects. Such an open canopy also facilitates better coverage with chemical sprays. In red varieties, there are increases in phenolics and tannins that contribute to flavour and complexity of the wine. Some remaining concerns include the development of undesirable flavour compounds in some white varieties and possible reduction in vine productivity. Additional studies are underway in Washington, United States of America, and at other North American locations in order to address these problems.

References

Araujo, F., Williams, L.E., Grimes, D.W. & Matthews, M.A. 1995a. A comparative study of young `Thompson Seedless' grapevines under drip and furrow irrigation. I. Root and soil water distributions. Scientia Horticulturae 60: 235-249.

Araujo, F., Williams, L.E., Grimes, D.W. & Matthews, M.A. 1995b. A comparative study of young `Thompson Seedless' grapevines under drip and furrow irrigation. II. Growth, water use efficiency and nitrogen partitioning. Scientia Horticulturae 60: 251-265.

Bravdo, B.A. & Hepner, Y. 1987. Irrigation management and fertigation to optimize grape composition and vine performance. Acta Horticulturae 206: 49-67.

Bucks, D.A., French, O.F., Nakayama, F.S. & Fangmeier, D.D. 1985. Trickle irrigation management for grape production. Drip/Trickle Irrigation In Action 1: 204-211.

Christensen, P.L. 1962. Boron deficiency in vineyards in Fresno county. Cooperative Extension Bulletin. Davis, California, United States of America, University of California.

Cline, R.A., Fisher, K.H. & Bradt, O.A. 1985. The effects of trickle irrigation and training system on the performance of Concord grapes. Drip/Trickle Irrigation in Action 1: 220-230.

During, H. 1986. ABA and water stress in grapevines. Acta Horticulturae 179: 413-420.

Evans, R.G., Spayd S.E., Wample R.L., Kroeger M.W. & Mahan M.O. 1993. Water use of Vitis vinifera grapes in Washington. Agricultural Water Management 23: 109-124.

Falcetti, M., Stringari, G., Bogoni, M. & Scienza, A. 1995. Relationships among pedo-climatic conditions, plant available water and nutritional status of grapevines. Acta Horticulturae 383: 289-297.

FAO. 1977. Guidelines for predicting crop water requirements by J. Doorenbos & W.O. Pruitt. Irrigation and Drainage Paper 24. Rome.

Fourie, J.C. 1988. Herbigation in a vineyard through a microjet irrigation system. South African Journal of Enology and Viticulture 9: 37-42.

Goldberg, S.D., Rinot, M. & Karu, N. 1971. Effect of trickle irrigation intervals on distribution and utilization of soil moisture in a vineyard. Soil Science Society of American Proceedings 35: 127-130.

Goodwin, I. & P. Jerie. 1992. Regulated deficit irrigation: from concept to practice. Wine Industry Journal 7: 258-261.

Grimes, D.W. & Williams L.E. 1990. Irrigation effects on plant water relations and productivity of Thompson Seedless grapevines. Crop Science 30: 255-260.

Hardie, W.J. & Considine, J.A. 1976. Response of grapes to water deficit stress in particular stages of development. American Journal of Enology and Viticulture 27: 55-61.

Mullins, M.G., Bouquet, A., & Williams, L.E. 1992. Biology of the grapevine. Cambridge, United Kingdom, Press Syndicate of the University of Cambridge.

Nadal, M. & Arola, L. 1995. Effects of limited irrigation on the composition of must and wine of Cabernet Sauvignon under semi-arid conditions. Vitis 34: 151-154.

Neja, R.A., Wildman, W.E., Ayers, R.S., & Kasimatis, A.N. 1977. Grapevine response to irrigation and trellis treatments in the Salinas Valley. American Journal of Enology and Viticulture 28: 16-26.

Peacock, W.L., Rolston, D.E., Aljibury, F.K. & Rauschlolb, R.S. 1977. Evaluating drip, flood, and sprinkler irrigation of wine grapes. American Journal of Enology and Viticulture 28: 193-195.

Smart, RE. & Coombe, B.G. 1983. Water relations of grapevines. In: T.T. Kozlowski ed. Water deficits and plant growth, Vol. 7. New York, United States of America, Academic Press.

Williams, L.E., Dokoozlian, N.K. & Wample, R.L. 1994. Grape. In: B. Schaffer & P.C. Andersen eds. Handbook of environmental physiology of fruit crops, Volume 1, Temperate crops, Boca Raton, Florida, United States of America, CRC Press.

Williams, L.E. & Matthews, M.A. 1990. Grapevine. In: B.A. Stewart and D.R. Nielsen eds. Irrigation of agricultural crops, agronomy Monograph No. 30, Madison, Wisconsin, United States of America, ASA-CSSA-SSSA.