|Most of South Africa’s wine is produced in the Western Cape, which has a Mediterranean climate where rainfall mainly occurs during winter, when grapevines require little water.
Since rainfall is infrequent and inadequate during the hot, dry summers, about 80% of the 110 200 ha of wine grape vineyards (Sawis, 2004) require irrigation. The frequency of irrigations varies from low, i.e. up to six per season in the cooler Coastal region, to intermediate where vineyards require irrigation at intervals between seven and 14 days. Intermediate frequency is mostly used in the warmer, drier areas such as the West Coast, Breede River valley and the Little Karoo. High frequency irrigation, i.e. where irrigation is required at intervals of less than seven days, is normally required on shallow, sandy soils, particularly when drip irrigation is used. Although the semi-arid Lower Orange River region is a summer rainfall area, the annual precipitation is less than 200 mm, which is totally inadequate to sustain viticulture. Consequently, intermediate to high frequency irrigation is needed in summer. Since South Africa is a relatively dry country, the water resources that are available for irrigation are limited. Against this background, efficient water use for wine production is an important consideration from social and environmental perspectives.
The water relations of grapevines, and in particular the effects of water stress as indicated by low leaf water potential, are well documented and reviewed (Smart & Coombe, 1983; Williams et al., 1994). However, practical irrigation strategies for specific combinations of climatic conditions, soil types, and vineyard practices with respect to optimum wine quality, are not so well documented. Water deficits can influence grape composition in more than one way. Plant water stress resulting from water deficits can retard sugar accumulation, particularly when grapevines bear high crop loads. Water stress can also decrease total titratable acidity, which may have a negative effect on wine quality. Colour development of red grapes can be delayed by both severe water stress and excess water.
Over the past few decades, a number of field trials have been carried out by ARC Infruitec-Nietvoorbij to determine the effects of different irrigation strategies on wine quality. Practical irrigation strategies and grapevine response will be discussed in terms of leaf water potential, vegetative growth, berry size, yield and sensorial wine quality. Although restrictions are inherent in the sensorial quality assessment of wines made on a small scale, sensorial assessment is preferred as a measure for the success of a particular irrigation strategy, rather than indirect quality assessments based only on berry size or the concentration of chemical substances. Only data for the most contrasting treatments with respect to wine quality will be discussed. Since irrigation strategies differ between the regions mentioned above, grapevine response will be presented and discussed separately for the Coastal region and for the warmer, drier inland areas.
The Coastal region: The proximity of the Atlantic Ocean influences the atmospheric conditions that vineyards are subjected to in the Coastal region. Mean February temperatures, i.e. when most of the wine grapes ripen, decline at a rate of circ. 0.5C per 100 m increase in altitude, and increase by circ. 0.6C per 10 km increase in distance away from the ocean (Myburgh, 2005a). Due to these effects, mean February temperatures vary between circ. 19C and 25C over the region. Irrigation water is obtained from bore holes and from dams which are either filled by surface run-off, or water that is pumped from rivers during winter. Given that these water resources are limited, effective soil preparation before planting will increase the amount of winter rainfall that can be stored in the root zone. A survey-type study in the Upper Berg River valley of the Coastal region, which included 103 Chenin blanc vineyards, showed that approximately 75% of them were either dry land (i.e. rain-fed) or received a maximum of two irrigations during the season as indicated in Table 1 (Van Zyl & Van Huyssteen, 1983). Since the inception of the Theewaterskloof scheme, a large number of previously dry land vineyards are now being irrigated, particularly in the Stellenbosch region.
If normal rainfall occurs during the preceding winter, the first irrigation is required when berries are pea sized (November). This is followed by irrigation at veraison (January) and post harvest (March), respectively, in vineyards with well-developed, deep root systems in clay and loamy soils. Under such conditions, low frequency irrigation events can be scheduled according to the phenological stages of the grapevine. Where available water is limited in gravelly or sandy soil, or where insufficient soil preparation restricts root depth, vineyards need to be irrigated before berries reach pea size. Irrigation will also be required more frequently during the rest of the growing season. Although post harvest irrigation might not have direct effects on yield or wine quality, it is required to facilitate the nutrient uptake which is needed for the synthesis and accumulation of reserves in the permanent structure of grapevines during autumn. Grapevines depend on these reserves to sustain growth in the following spring when their roots are still inactive.
The survey in the Upper Berg River valley showed that yields still respond favourably to irrigation, although vineyards are irrigated at low frequencies (Van Zyl & Van Huyssteen, 1983). Four to six irrigation events almost doubled the yield as compared with non-irrigated vineyards (Table 1). However, the highest yield increase per irrigation was obtained with up to three irrigations. At the time that the survey was done, most vineyards were either irrigated by means of overhead or micro-sprinklers which wetted the full soil surface. Normally these full surface irrigations varied between 50 mm and 100 mm per application, depending on soil water storage capacity and root depth. Since the wine industry was primarily production driven when the data was collected, no attention was paid to parameters of wine quality.
Some of the earliest research results showed that the number and timing of low frequency irrigation events have definite effects on Chenin blanc yield and wine quality, but that the positive effects on wine quality were not consistent over seasons (Van Zyl & Weber, 1977). More recently, a study was initiated in a Merlot vineyard near Wellington to compare the partial root zone drying (PRD) irrigation strategy against conventional low frequency drip irrigation as indicated in Table 2. To avoid the high costs of the double irrigation lines required for PRD irrigation, the latter was applied through a single sub surface line in the centre of the work rows. The PRD effect was obtained by irrigating twice a week in alternative work rows. Irrigations were switched approximately every 14 days to the work rows that had previously been allowed to dry out. To enable comparison to be made between treatments, conventional low frequency irrigations were also applied sub surface in the work row. A non-irrigated treatment was included to serve as a control. Although the amount of water required for the PRD strategy was comparable to the amount required for the four irrigations, water stress in the grapevines which received four irrigations was consistently higher compared to PRD grapevines after the irrigation season started in mid November (data not shown). The effects of the different irrigation strategies were also reflected in the leaf water potential measured on the day before the grapes were harvested (Table 2). The high levels of water stress in grapevines in the drier treatments did not seem to cause any severe delay in sugar accumulation, and grapes of all treatments could be harvested on the same day. The fact that less water stress did not increase growth accordingly indicated that the PRD strategy curbed vegetative growth to some extent (Table 2). Furthermore, less water stress did not increase berry size and yield compared to four irrigations. However, berry size and yield of the grapevines which received four irrigations as well as those of the PRD strategy were higher compared to the dry land grapevines, and those which received only two irrigations (Table 2). During the 2003/04 season two irrigations tended to improve wine colour, berry aroma intensity and overall wine quality (Table 3). Visual observation revealed that the dry land grapevines were seriously damaged during a heat wave that occurred on 9 February 2004. In the more moderate 2004/05 season, when no severe heat waves occurred during ripening, dry land conditions resulted in better wine quality compared to the PRD strategy. As in the first season, the more conventional irrigation strategy of only two irrigations had a positive effect on berry aroma and wine quality compared to the PRD strategy. The foregoing confirmed that limited irrigation in the Coastal region will increase yields, and that it is unlikely to have any negative effects on wine quality. Furthermore, these results provide evidence that judicious irrigation can improve wine quality, depending on the climatic conditions during a specific season. The results also showed how irrigation can reduce variation in wine quality from season to season. Therefore, it may be accepted that low frequency irrigation will increase the profitability of vineyards in the Coastal region.
When more water is available for irrigation, growers are tempted to irrigate more frequently, particularly during the berry ripening period, and especially with drip irrigation. This aspect was studied over three seasons in a micro-sprinkler irrigated Sauvignon blanc vineyard near Stellenbosch (Myburgh, 2005b & 2006). At harvest, water stress in grapevines was less where additional irrigation was applied at veraison and during ripening compared to a single irrigation at pea size (Table 4). Compared to the latter treatment, where almost all the available water had been depleted when the grapes were harvested, an additional irrigation at veraison resulted in higher berry mass. A third irrigation applied one week before harvest also increased berry mass, but, as in the case of the two irrigations, the higher berry mass did not result in a higher yield. Although wine fullness tended to be higher where irrigation was applied shortly before harvest, it did not seem to have any effect on overall wine quality compared to the single irrigation that was applied at pea size. The lack of positive wine quality response was probably due to the fact that the dominant aroma of the fuller wine was, according to the tasting panels, not the expected vegetative aroma for Sauvignon blanc. This trend was consistent over the three seasons. During one season, rain that occurred shortly before harvest, as well as wetter soil conditions caused by irrigation during berry ripening, induced considerably more Botrytis cinerea infection. Hence, additional irrigation during the ripening period will not necessarily increase yield, but may increase the risks of disease infection and reduced wine quality. Where available water in the root zone is limited due to sandy soil or shallow root systems, irrigation during ripening may be necessary to support the ripening processes, particularly sugar accumulation. However, the last irrigation should still be applied at such a time that circ. 90% of the available water will be depleted at harvest.
Warmer inland regions: In contrast to the Coastal region, mean February temperatures for the inland regions vary from circ. 22C to 24C for the Breede River valley and Little Karoo to more than 25C in the Lower Orange River region. Due to the warmer climate and adequate water from extensive irrigation schemes, as well as the drive for high yields, vineyards in these regions were traditionally irrigated at relatively high frequencies. However, the increasing demand for high quality wines by export markets could provide an incentive for growers to irrigate more judiciously – at the cost of higher yields.
To obtain guidelines for moderate frequency irrigation, the effects of different levels of plant available water (PAW) depletion on yield and wine quality were determined in a number of field trials. In these studies, PAW was defined as the soil water that is retained between field capacity and -0.1 MPa matric potential. For the soils in which the trials were carried out, field capacity is at matric potentials of -0.005 MPa and -0.01 MPa for sandy and heavier soils, respectively. A high level of depletion, e.g. 80%, implies that the soil will be allowed to become considerably drier between irrigations in comparison to irrigation at a lower level of PAW depletion, e.g. 40%.
In the Breede River valley, irrigation at 75% PAW depletion throughout the season reduced berry mass of micro-sprinkler irrigated Colombar grapevines in loamy soil compared to 30% and 50% PAW depletion (Van Zyl, 1984a). Drier soil conditions reduced shoot growth, and irrigation at 75% PAW depletion throughout the season tended to reduce yield, also increasing overall wine quality (Table 5). These trends were relatively consistent over six seasons. Where grapevines were irrigated at 75% PAW depletion after flowering, i.e. during the first stages of berry development, followed by irrigation at 30% depletion during the rest of the season, berries were also smaller compared to irrigation at 30% depletion maintained throughout the season (Van Zyl, 1984a). Although berries were smaller, wine quality was not improved by early season water deficits. Similarly, wine quality only tended to be higher where smaller berries were obtained when Colombar was irrigated by means of micro-sprinklers at different frequencies in the hot, semi-arid Lower Orange River region (Table 6). It must be noted that the lower irrigation frequencies, i.e. 21 to 28 day intervals, did not have any negative effects on yield and wine quality in the latter region where irrigations are normally applied at 14 day intervals. Furthermore, it is of particular interest that in both studies with Colombar, drier soil conditions tended to improve wine quality of this fertile cultivar bearing crop loads in excess of 30 t/ha in these warmer climatic conditions.
More recently, the effects of water deficits on wine quality in Pinotage and Sauvignon blanc were determined over three seasons in the Breede River valley. Early season, late season and continued water deficits throughout the growing season were compared to irrigation at 50% PAW depletion. Water deficits were induced by allowing 80% of PAW to be depleted before irrigation was applied. Irrigation was required at approximately weekly and 14 day intervals to maintain the respective 50% and 80% PAW depletion levels. Irrigation applied at 80% PAW depletion throughout the season did not reduce vegetative growth of Pinotage trained onto a two-tier trellis compared to irrigation at 50% depletion, but reduced berry mass (Table 7). The lower berry mass contributed to the circ. 11% lower yield obtained with 80% PAW depletion compared to 50% depletion.
Anthocyanin concentrations in Pinotage wines only tended to be higher where irrigation was applied at 80% PAW depletion compared to 50% depletion (Table 8). However, irrigation at 80% PAW depletion produced fuller wines with more intense cultivar aroma and higher overall wine quality compared to irrigation at 50% depletion. Similar to the results obtained for Colombar (Van Zyl, 1984b), water deficits during the early stages of berry development reduced berry mass, but did not improve wine quality compared to irrigation at 80% PAW depletion throughout the season (data not shown). Under the same conditions, grapevines on a single cordon vertical trellis produced wine of similar quality, but lower yields compared to those on the two-tier trellis (data not shown).
Although significant differences in water stress between treatments occurred in Sauvignon blanc grapevines that were trained onto a single cordon vertical trellis (data not shown), irrigation at 50% and 80% PAW depletion, respectively had no effect on cane mass, berry mass and yield (Table 9). This suggested that Sauvignon blanc must be subjected to severe water deficits in order to reduce berry mass significantly, as was reported for low frequency irrigated vineyards near Stellenbosch (Myburgh, 2005b). Under the given conditions, Sauvignon blanc wine quality also seemed to be insensitive to water stress, which is in agreement with the results of the study at Stellenbosch (Myburgh, 2006). As in the case of Pinotage, vegetative growth and yield was higher on a two-tier trellis, but wine quality was not affected. For both cultivars, circ. 500 mm were required to maintain 80% PAW depletion compared to 750 mm required for 50% depletion, irrespective of trellis system. This is a significant finding regarding water resource utilisation, since more efficient water use will become an increasingly important consideration in future when it comes to water use license allocations. Using a system such as the two-tier trellis will enable growers to produce more grapes per unit land where arable land is limited.
Irrigation strategies, as applied in the various field trials, seemed to have limited effects on wine quality, particularly in the case of white cultivars. A possible reason might be that in most experiments, the applied treatments were relatively dry, i.e. PAW depletion levels were between 40% and 90%, and that wine quality response in this depletion range is relatively small. A comparison between wetter soil conditions, e.g. where irrigation is applied at 20% PAW depletion, and the levels that were maintained in most of the field trials, might have caused more prominent wine quality differences. Since wetter soil conditions are more likely to reduce wine quality, such treatments will not make a meaningful contribution if they are included in future research projects. However, under the given conditions, it is evident that increased irrigation will increase grapevine yield, but that the higher yields are not necessarily related to inferior wine quality if available water depletion is maintained at circ. 50%. This is an important consideration for producing above-average wines in warmer regions. The PRD strategy could be useful to curb vegetative growth of vigorous cultivars without reducing yield, but will not necessarily improve wine quality. Furthermore, it is clear that water stress levels, which induce a combination of lower vigour and smaller berries, can almost invariably be related to a trend towards higher wine quality. This positive trend also seems to be applicable to cultivars such as Merlot and Sauvignon blanc which are regarded as being sensitive to water stress. Due to the complexity of the factors that influence wine quality, it is not possible to single out a factor that can explain the differences and trends in wine quality. This means that, e.g. the smallest berries or highest anthocyanin concentration cannot be accepted as a guarantee for optimum wine quality under a given set of conditions. Considering the response of all the relevant variables to irrigation, the likelihood of obtaining and maintaining optimum wine quality will increase when irrigations are applied when 80% to 90% of PAW is depleted.
Agricultural Research Council for infrastructure and other resources, Winetech for partial funding and Soil Science Staff at ARC Infruitec-Nietvoorbij for technical support.
For further information contact Philip Myburgh at firstname.lastname@example.org.
Myburgh, P.A. 2005a. Effect of altitude and distance from the Atlantic Ocean on mean February temperatures in the Western Cape Coastal region. Wineland, August 2005, 84-86.
Myburgh, P.A. 2005b. Water status, vegetative growth and yield of Vitis vinifera L. cvs. Sauvignon blanc and Chenin blanc in response to timing of irrigation during berry ripening in the Coastal region of South Africa. S. Afr. J. Enol. Vitic. 2, 59-67.
Myburgh, P.A. 2006. Juice and wine quality responses of Vitis vinifera L. cvs. Sauvignon blanc and Chenin blanc in response to timing of irrigation during berry ripening in the Coastal region of South Africa. S. Afr. J. Enol. Vitic. (accepted for publication).
SAWIS, 2004. South African Wine Industry Information & Systems. P.O. Box 238, 7620 Paarl, South Africa.
Smart, R.E. and Coombe, B.G. 1983. Water relations of grapevines. In: Kozlowski, T.T. (ed.). Water deficits and plant growth, Vol. VII. Academic Press, New York. 137-196.
Van Zyl, J.L. 1984a. Interrelationships among soil water regime, irrigation and water stress in the grapevine (Vitis vinifera L.). Thesis, University of Stellenbosch, Private Bag X1, 7602 Matieland, South Africa.
Van Zyl, J.L. 1984b. Response of Colombar grapevines to irrigation as regards quality aspects and growth. S. Afr. J. Enol. Vitic. 5, 19-28.
Van Zyl, J.L. and Weber, H.W. 1977. Irrigation of Chenin blanc in the Stellenbosch area within the framework of the climate-soil-water-plant continuum. Proc. Int. Symp. quality vintage, Cape Town, South Africa. 331-349.
Van Zyl, J.L. and Van Huyssteen, L. 1983. Soil and water management for optimum grape yield and quality under conditions of limited irrigation. Proc. 5th Austr, Wine Indus. Tech. Conf., 29 November – 1 December 1983, Perth. 25-66.
Williams, L.E.; Dokoozlian, N.K. and Wample, E.R. 1994. Temperate crops. In: Schaffer, B. & Anderson, P.C. (eds). Handbook of environmental physiology of fruit crops, Vol 1, CRC Press, Boca Raton. 85-133.