Seasonal fluctuations in yield, grape composition and wine attributes, largely driven by variable climatic conditions, are major challenges for the wine industry aiming to meet consumer expectations for consistent supply, wine style and product quality. This paper will address known causes for this variability and identify management techniques, together with their limitations, that offer potential to modulate these responses.
Results will be presented from research studies, conducted over a number of seasons, which link vineyard management practices with fruit composition and wine assessments. They show that there is potential to develop integrated systems to stabilise yield, fruit composition and wine quality attributes across seasons. These techniques include lighter pruning, deficit irrigation techniques and adoption of low-moderate vigour rootstocks which lead to reduced shoot vigour and the development of open canopies and small bunches with small berries, with enhanced colour, phenolics and sensory appeal; application of mechanical and chemical crop thinning techniques for yield stabilisation and promotion of early ripeness, colour and flavour development.
However, the results also show that variability between seasons in many cases is much larger than can be achieved by modifying management practices. Hence, opportunities to use management practices to completely reduce the seasonal variability, particularly with respect to grape composition and wine attributes, factors largely affected by climatic conditions during berry development, may be limited. There exists significant potential in the longer term to use new varieties and rootstocks better adapted to variable and changing climatic conditions.
It can be concluded that vineyard practices can be modified to stabilise yield, grape composition and wine attributes to varying degrees and hence, minimise the impact of variability in climatic conditions from season to season. In the longer term, the adoption of new varieties together with new rootstocks adapted to higher temperatures and limited water supply will assist the wine industry to cope with impacts of climate variability and change and address ever changing consumer expectations. A total systems approach to vineyard management offers potential to modulate seasonal fluctuations in yield, grape composition and wine quality attributes with significant benefit for an industry.
Seasonal fluctuations in yield, grape composition and wine attributes, largely driven by variable climatic conditions, are a major challenge for the Australian wine industry aiming to meet consumer expectations for consistent supply, wine style and product quality. In recent seasons, problems related to limited water supply and impacts of drought have compounded these issues. Furthermore, to be economically sustainable, the industry must also continue to improve production efficiency (i.e. higher yields for efficient use of natural resources) with reduced inputs (i.e. labour, fuel, water and chemicals). The industry also needs to address the issue of high alcohol content of many Australian wines which has been steadily increasing over the last decade (Clingeleffer 2008). From a grape production perspective, the latter challenge implies that wine quality attributes (i.e. mouthfeel, aroma, flavour and colour) are maintained from fruit harvested with lower sugar levels. It has become apparent that the industry's capability to address these challenges through adoption of vine management practices is limited by the dominant effects of season to season variation, not only on productivity, but also on berry growth and development, fruit composition and final wine quality attributes.
This overview will address known causes for the seasonal variability and identify management techniques, together with their limitations, that offer potential to modulate these responses. Results will be presented from research studies, conducted over a number of seasons, which link vineyard management practices with fruit composition and wine assessments. They show that there is potential to develop integrated systems to stabilise yield, fruit composition and wine quality attributes across seasons in a low input context. The results also highlight that knowledge of the processes controlling crop development and berry composition in relation to impacts of environmental cues and imposition of vineyard management practices is limited.
Long-term trends in yield and composition
The long-term yield trend for Australian Sultana grown in the warm irrigated regions is shown in Figure 1. Over time there has been a 2.5-fold increase in productivity which can be linked to the combined benefits from the adoption of improved clones, high vigour and nematode tolerant rootstocks such as Ramsey, lighter pruning, larger trellises and improved irrigation practices (Clingeleffer 1994). The data also highlight the significant variation from season to season which can be linked to environmental influences on shoot fruitfulness in the season prior to harvest (Baldwin 1964) and crop development in the harvest season. For Sultana growing under Australian conditions, Baldwin (1964) reported a 20-day ‘sensitive period’ from mid-November to early December, during which bright sunshine and temperature were most significant in the determination of bud fruitfulness. Similar studies have not been undertaken to assess environmental influences on fruitfulness of winegrapes grown under Australian conditions although May and Cellier (1973), in studies across ten cultivars suggested that similar processes do occur. Similar sensitive periods have been reported for wine grapes growing under cooler conditions in the Northern hemisphere including Riesling (Alleweldt 1963) and Chardonnay (MacGregor 2002). In the case of Chardonnay, the reported fourfold difference in shoot fruitfulness, ranging from 0.3–1.3 bunches per shoot was associated with differences in average temperature during a 20-day sensitive period commencing 10 days before flowering (MacGregor 2002). In controlled environment studies, Buttrose (1970) showed that inflorescence initiation of Shiraz and Riesling was severely reduced below 20°C. Further analysis of the long-term Sultana yield data by Petrie and Clingeleffer (unpublished), has linked yield to temperature during critical phases of crop development. In the preceding initiation season, these include positive responses during anlagen induction in September (i.e. >18°C) and the sensitive period around flowering in November–December, and a negative response in the pre-veraison, late December period (>26°C). During crop development in the harvest season, they found a positive association with budburst temperature (>15°C) and a negative effect of temperature post-veraison (>27°C). While the causes for season to season responses are not fully understood, they may be linked to factors influencing primordia size, e.g. carbohydrate levels in buds and dormant canes as reported for Sultana (Antcliff and Webster 1955, Sommer et al. 2000), frequency of frost during budburst and effects of stress on photosynthesis and subsequent effects on berry growth and sugar accumulation.
Long-term yield trends for Cabernet Sauvignon, growing in three very different Australian regions (i.e. Sunraysia, Barossa and Coonawarra) are presented in Figure 2. The mean January temperatures and growing degree days (HDD°C) for the Sunraysia (or Murray Darling region), Barossa and Coonawarra regions are 23.9°C, 21.4°C and 19.6°C and 2244 HDD°C, 1715 HDD°C, 1432 HDD°C, respectively (Dry and Smart 1988). As with Sultana, the trend show a positive yield trend over time, particularly in Sunraysia and the Barossa Valley, which can be linked to lighter pruning, adoption of improved clones and rootstocks and improved irrigation practices (Clingeleffer et al. 1997). However, in all three regions, this trend has not continued since the early 1990s, possibly because of influences of wineries on targeted yields. The results also show very large variation in yield from season to season. Clingeleffer et al. (1997) established a link with low temperatures in the Coonawarra region which showed that 27% of the seasonal variation in yield could be accounted for by the total number of days when frost occurred in the pre- and post-budburst period.
Overall, the links between climatic factors and performance of winegrapes grown under Australian conditions are poorly established. Analyses of longer term data sets, similar to that undertaken by Baldwin (1964) and Petrie and Clingeleffer (unpublished, see above) would help identify critical periods during crop development which then could be targeted for further research. Analyses of yield components in trials in key winegrape regions have shown that invariably bunch number is the key driver of yield (Clingeleffer et al. 1997, Martin et al. 2002), accounting for 50–80% of seasonal yield variability with berries per bunch (a function of flower number and fruit set) accounting for 10–20% of the variability. Hence it is important that more research is undertaken to link environmental cues with effects on bunch number. This must include not only research on fruit bud induction and initiation as previously described for Sultana but also account for other factors during crop development in the harvest season. The period during which budburst occurs appears to be a critical period which is poorly understood. For example, by comparing dormant bud fruitfulness and shoot fruitfulness, loss of inflorescences at budburst has been observed in Chardonnay (Welsh et al. 2005) but not in Shiraz or Cabernet Sauvignon. Pouget (1981) reported an increase in shoot fruitfulness with higher temperatures at budburst for Cabernet Sauvignon and Merlot in pot experiments. Petrie and Clingeleffer (2005) in a field study with Chardonnay, found a strong negative correlation between temperature and flower number (R2 = 0.81). Furthermore, significant losses of bunches between inflorescence counts in spring and bunch counts at harvest have been recorded. For example, in Sunmuscat, this phenomenon has been linked to pruning severity with losses ranging from 17% for severely pruned vines (6 canes) to 39% for lightly pruned vines (15 canes) (Clingeleffer and Tarr, unpublished data). Such results suggest that bunch retention is an assimilate driven, adaptive process. Other factors contributing to seasonal variation in bunch number include losses associated with necrosis of the primary bud which occurs in Shiraz (Dry and Coombe 1994) and complete bunch stem necrosis (Stellwaag-Kittler 1983, Holzapfel and Coombe 1995).
Significant changes in fruit composition and wine quality attributes over time should also be of concern to an industry seeking economic sustainability. For example, Clingeleffer and Lo Iacono (2005) reported for Cabernet Sauvignon grown in a warm irrigated region an increase in must pH (3.1–3.8, R2 = 0.65) and a decrease in wine titratable acidity, ionised anthocyanins, total anthocyanins and phenolics over a 13-year period (1989–2001). These effects could not be linked to harvest date, berry weight or sugar level at harvest. A similar result has been found for Shiraz grown in Coonawarra with a significant increase in pH (3.3–3.7, R2 = 0.82) over a 10-year period, 1992–2001 (Clingeleffer, unpublished data). While one can only speculate as to the causes for the large changes in composition over time, effects of vine age, depletion of soil nutrients (e.g. divalent ions, Mg+ and Ca+), subtle changes in vineyard management and potential effects of climate change require further investigation.
Light pruning systems
Research over 30 years has shown that traditional regulation of bud numbers by hand pruning is unnecessary and may contribute to low wine quality generally associated with development of shaded, tight bunches with large berries and difficulties in the control of pests and diseases (Clingeleffer et al. 1999, 2000, Clingeleffer 2000). Beneficial attributes associated with lighter pruning systems include development of vines with small, well-exposed bunches of small berries, spread over a large canopy surface which lead to good disease control and improved berry composition, provided adequate sugar levels are reached (Clingeleffer 2000, Clingeleffer et al. 2000). Compared with severe forms of pruning, minimally pruned vines generally produce juice with better organic acid composition, expressed as a superior tartrate to malate ratio, better wine colour and higher phenolics (Clingeleffer 1993, Clingeleffer 2000). Similarly, for Riesling and other white varieties grown in a very cool climate in Germany, minimal pruned vines compared with those grown on a standard vertical shoot positioned system, had higher yields (25–75%), looser clusters and smaller berries with higher concentrations of aromatic precursors. This was reflected in enhanced sensory perceptions, despite having slightly lower levels of sugar (Schultz and Weyand 2005). Delays in veraison and harvest, often considered a negative attribute of lighter pruning systems, in particular minimal pruning, may indeed be of positive benefit in the context of climate change and production of lower alcohol wines, as positive fruit composition and wine quality attributes may be achieved at lower sugar levels.
The significant effects of both season and contrasting pruning treatments (spur, mechanically hedged and minimal pruning) on crop development, berry and must composition, and wine spectral properties for Shiraz grown in the warm climate, Sunraysia region (mean January temperature of 23.9°C, Dry and Smart 1988), are shown in Table 1. Compared with spur pruning, the hedged and more significantly the minimal pruned vines over the three seasons produced more bunches, shorter inflorescences with fewer branches and lower fruit set leading to development of smaller bunches with fewer smaller berries and modest increases in yield, i.e. 12 and 22%, respectively. Despite the large developmental responses to lighter pruning treatments, including berry size (i.e. 10 and 25% reductions for the hedged and minimal pruned treatments, respectively), the effects on berry composition and wine spectral properties were small, although generally significant. Although the minimal pruned vines were harvested at lower ripeness than the spur or hedged vines, the berries had the highest phenolics and produced wines with the highest colour density, ionised anthocyanins, and total phenolics. Compared with the spur pruned vines, hedged vines produced wines with higher colour density and ionised anthocyanins. However, the season to season variation in crop development, berry composition and wine spectral properties were large, as shown for hedged vines, the standard vineyard treatment. Between seasons there were major differences in bunch number (threefold), flowers per inflorescence (2.3-fold), fruit set (2.7-fold) and yield (2.1-fold) with significant differences between seasons in inflorescence length (35%), branches per inflorescence (27%), flowers per inflorescence (28%), berries per bunch (37%), bunch weight (60%), bunch length (35%) and berry weight (28%). These results highlight the plastic nature of the vine and its extraordinary ability to compensate during crop development in the harvest season for low fruitfulness, as occurred in season 2003. As discussed previously, the environmental cues contributing to such large season to season variability in fruitfulness are poorly understood. It is interesting to note the low fruitfulness, as determined from bunch number in season 2003 was accompanied by formation of short inflorescences with fewer branches and flowers. In contrast to the limited pruning responses for berry composition and wine spectral properties, seasonal effects were much larger with significant differences in berry anthocyanins (28%), berry phenolics (13%), titratable acidity (52%), pH (10%), wine colour density (31%), ionised anthocyanins (22%), total anthocyanins (5%) and total phenolics (13%). The effect of season on berry composition and spectral properties could not be linked to yield or fruit ripeness (e.g. season 2004 had the highest yields and lowest sugar but the highest wine colour density and ionised anthocyanins) and is most likely associated with climatic conditions during berry development. For example, high temperature around veraison has been linked to lower wine quality (i.e. vintage scores) across four Australian wine regions (Soar et al. 2008).
Table 1. Means of yield components, berry composition and wine spectral properties of Shiraz managed under three pruning treatments (spur, hedge, minimal) over three seasons (2002–2004), and for hedged vines each season.
Superscripts are used to identify means which are significantly different (P = 0.05). For each variable the ratio of the highest to lowest value is included. Data adapted from Clingeleffer et al. (2005) and Petrie and Clingeleffer, unpublished data. †Mechanically hedged treatment only.
Bunch number per metre
Inflorescence length (mm)
Branches per inflorescence
Flowers per inflorescence
% fruit set
Berries per bunch
Bunch weight (g)
Bunch length (mm)
Berry weight (g)
Berry anthocyanins (mgg–1)
Berry phenolics (mgg–1)
Must sugar (0Brix)
Must titratable acidity (gL–1)
Wine colour density (a.u.)
Ionised anthocyanins (mgL–1)
Wine total anthocyanins (mgL–1)
Wine total phenolics (a.u.)
A further example of the benefits from the adoption of light pruning of Cabernet Sauvignon, grown in the warm climate, Sunraysia region (mean January temperature of 23.9°C, Dry and Smart 1988) was demonstrated by Petrie et al. (2003a). Vines were trained under four pruning regimes including hand spur pruning, tight and loose mechanical hedging (cuts applied to give a pruned width of 0.4 m and 0.6 m, respectively) and minimal pruning. Yield was not affected by the treatments. However, as pruning severity decreased, bunch numbers increased (from 74 to 243), bunch weight decreased (from 68.8 to 23.7 g) and berry weight decreased (from 1.03 to 0.76 g). The fruit was harvested on the same day with moderate levels of total soluble solids for the spur and mechanical hedging treatments (i.e. between 23.0 and 23.6°Brix), while ripeness was delayed with minimal pruning (21.9°Brix). Spur pruning decreased anthocyanin concentrations in berry homogenates compared with the other treatments (i.e. 0.55, 0.67, 0.84 and 0.68 mg/g for the spur, 0.4 and 0.6 m hedging treatments and minimal pruning, respectively). The lighter mechanical pruning (0.6 m) had significantly higher levels of berry anthocyanins than the 0.4-m wide mechanical pruning, demonstrating how a subtle difference in mechanical pruning can cause a substantial difference in fruit quality. Despite the lower ripeness from minimal pruned vines, anthocyanin levels in the fruit were similar to the 0.4-m hedge treatment and higher than the spur pruned vines indicating potential for production of low alcohol wine without compromising wine attributes, e.g. colour.
For Cabernet Sauvignon grown in the a cool climate Mornington region (mean January temperature of 19.0°C), three pruning systems were compared (i.e. traditional 6-cane system with vertical shoot positioning, hanging canes arising on a 1.8-m high cordon and minimal pruning established on the 1.8-m cordon (Clingeleffer 1993)). Compared with cane pruning which had a yield of 12.7 t/ha, a berry weight of 0.95 g and berry ripeness of 21.8°Brix, both the hanging cane and minimal pruning treatments produced higher yields (19.4 and 18.9 t/ha, respectively), smaller berries (0.74 g) but delayed ripeness (20.4 and 21.3°Brix, respectively). Despite the lower levels of soluble solids, fruit from both the hanging cane and minimal pruning treatments had higher tartrate to malate ratios and produced wines with higher colour density (i.e. 20.1 and 19.1 a.u., respectively, compared with 15.3 a.u. for cane pruning), higher total anthocyanins (1030, 962 mg/L, respectively, compared with 850 mg/L for cane pruning) and higher total phenolics (68, and 64 a.u., respectively, compared with 54 a.u. for cane pruning). Wines from the higher yielding, lighter pruned treatments were also preferred in sensory assessments by industry taste panels, which may have been linked to changes in wine aroma and flavour attributes as found in detailed hand pruning studies (Chapman et al. 2004). For Cabernet Sauvignon grown in the Napa Valley in California, Chapman et al. (2004) found that wines produced from lighter pruning treatments had increased intensity of ‘fruity’ attributes and reduced ‘vegetal’ character compared with wines from severely pruned treatments when fruit was harvested at similar maturities.
Post-fruit set deficit irrigation treatments have been widely adopted by growers with the aim to improve fruit and wine composition. However, the potential for such treatments to modulate season to season variability in yield, fruit composition and wine quality attributes has not been reported. In one study, two deficit irrigation treatments (ie. post-set regulated deficit irrigation (RDI) and a prolonged deficit (PD) applied up to veraison in addition to the post-set RDI treatment) were compared with an unstressed control treatment for own rooted, Cabernet Sauvignon over 3 years (Cooley et al. 2005). Compared with the unstressed control treatment, the RDI and PD treatments produced smaller canopies and lower yields (i.e. 7 and 15% lower than the control, respectively) due largely to the development of smaller berries (i.e. 9 and 12% lower, respectively). Over the three seasons, soluble solids of all treatments were similar at harvest (i.e. around 24°Brix). Spectral properties of wines from the RDI and PD treatments were enhanced with increasing levels of stress (i.e. increases in colour density by 10 and 27%, total anthocyanins by 14 and 30% and total phenolics by 12 and 14%, respectively). Intensive sampling during ripening confirmed that there were no significant effects of the water stress treatments on ripening. However, with the RDI and PD treatments, berry anthocyanin concentrations were generally higher at lower juice soluble solids (i.e. 20–22.5°Brix in mid-February) with mean values of 1.6 and 1.9 mg/g, respectively, compared with 1.4 mg/g for the control. Seasonal variability in berry anthocyanins was reduced with both deficit treatments, particularly the PD treatment compared with the control. The results indicate that deficit irrigation treatments have potential to both modulate effects of season on berry composition, e.g. anthocyanins, and also enhance wine spectral properties at lower levels of soluble solids. More recent studies from the same trial site have confirmed that the above-mentioned trends have been maintained over six seasons and have also shown an increase in wine tannin concentration with increasing levels of water stress (Edwards and Clingeleffer, unpublished).
Integration of light pruning systems and irrigation practices
The combined benefits of integrating light pruning treatments with a reduction in application of irrigation water using partial rootzone drying (PRD) was assessed for Shiraz grafted on Schwarzmann rootstock and grown in a warm irrigated region. A PRD treatment (2.5 ML/ha applied each season) was compared with control deficit irrigation (5 ML/ha applied each season) over three pruning treatments (i.e. spur, mechanical hedged and minimal pruned) for three seasons, 2001–2003 (Clingeleffer et al. 2005, Ashley et al. 2006). Compared with the control, across all pruning treatments the PRD treatment reduced yield by 19%, because of the development of smaller bunches with smaller berries, and produced a small, but significant delay in ripeness (i.e. 23.8 compared with 24.0°Brix for the control). There was no significant effect of PRD on berry composition (pH, titratable acidity, anthocyanin and phenolic concentrations). However, PRD wines had higher colour density, a brighter colour (hue) and higher total anthocyanins and phenolics compared with the control irrigation treatment.
Compared with the spur pruned vines, lighter pruning increased yield (i.e. by 22 and 25% for the hedged and minimal pruned treatments, respectively) because of development of more but smaller bunches (Clingeleffer et al. 2005, Ashley et al. 2006). Although the lighter pruning treatments had little effect on ripeness (23.7°Brix for both compared with 24.0°Brix for the spur treatment), they produced a significant impact on berry composition with lower pH, higher titratable acidity, higher phenolic concentration but no effect on anthocyanin concentrations. Compared with spur pruning, wines from lighter pruning treatments had increased colour density, higher levels of ionised anthocyanins and phenolics (minimal pruning treatment only) but similar levels of total anthocyanins.
In general, there were no interactions between the irrigation and pruning treatments (Ashley et al. 2006). Benefits from integration of both the PRD and lighter pruning treatments can be assessed from comparisons of the extreme treatments (i.e. spur pruned-control irrigation and minimal pruning-PRD treatments). Compared with the spur pruned-control irrigated treatment, the minimal pruned-PRD treatment had higher yields (9%), lower soluble solids (23.7 compared with 24.5°Brix) and produced wines with higher wine colour density (9.9 compared with 8.2 a.u), higher total anthocyanins (744 compared with 658 mg/L) and higher phenolics (65 compared with 56 a.u.). These results indicate that integration of light pruning and PRD treatments, not only increased water use efficiency (i.e. irrigation water use index measured as yield per irrigation water applied) and overcame the yield losses found with PRD, but also produced wines with enhanced spectral properties despite the lower ripeness at harvest.
Crop control techniques
Application of crop control techniques, other than by winter pruning, has potential to minimise seasonal yield fluctuations, manipulate canopy to fruit ratios and provide the winemaker with a consistent supply of fruit, produced to specifications of a given wine style. Crop adjustment after fruit set of minimal pruned Cabernet Sauvignon and Shiraz, either by mechanical thinning using mechanical harvesters or skirting, has been successfully applied in both cool and warm regions (Clingeleffer et al. 2000, 2002, Petrie et al. 2003b,c, Petrie and Clingeleffer 2006). In general, crop adjustment compared with the untreated control treatment, promoted earlier ripening and produced desirable changes in wine quality parameters (i.e. higher tartrate to malate ratio, higher levels of titratable acidity, improved spectral qualities and higher levels of phenolics). It is highly likely that fruit from the thinned treatments would be more suitable for wine production when harvested at lower sugar levels, based on detailed studies of fruit composition over the ripening period. In recent years, a range of mechanical and chemical techniques for control of cropping levels by removal of shoots, berries, inflorescences and bunches are being reported in the literature. For example, gibberellic acid treatments have been successfully applied to reduce crop load and subsequent fruitfulness of minimal pruned Riesling in Germany while maintaining the quality benefits associated with minimal pruning, i.e. reduced bunch rot and enhanced flavour and aroma characteristics (Weyand and Schultz 2006). Furthermore, pre-bloom mechanical defoliation of basal leaves has been successfully used to regulate yield of Sangiovese in Italy, by reducing fruit set (Intrieri et al. 2008).
In one such study with Shiraz grafted on Ramsey rootstock and grown in the warm irrigated region of Sunraysia, mechanical crop thinning was applied in early December, 3 weeks after the completion of fruit set (Clingeleffer et al. 2002). Fruit of the mechanically thinned treatment ripened faster than the unthinned control, reaching a moderate level of soluble solids (i.e. 22°Brix) almost 3 weeks earlier. Fruit for winemaking was harvested from the thinned treatment on 8th March and from the control on 23rd March. The thinning treatment reduced the yield by 36% (i.e. from 35 to 23 tha−1) because of the combined effects of bunch removal, lower berry numbers per bunch and development of smaller berries. Despite the similar ripeness at harvest (i.e. 22°Brix), compared with the unthinned control, the thinning treatment produced fruit with a significantly lower pH (i.e. 3.80 compared with 4.22), higher titratable acidity (i.e. 5.20 compared with 3.72 g/L) and higher anthocyanins (1.10 compared with 0.82 mg/g). Compared with the control, wines from the thinned treatment had improved spectral qualities (i.e. increases in colour density of 36%, ionised anthocyanins of 34%, total anthocyanins of 48% and phenolics of 46%) and received higher sensory scores when assessed by industry taste panels.
CSIRO has recently released three new low-moderate vigour rootstocks for larger scale industry evaluation. These rootstocks, Merbein 5489, Merbein 5512 and Merbein 6262 are described in Dry (2007). Pruning weights of Shiraz grafted on the new rootstocks were 2–3.5-fold lower than standard high vigour, nematode tolerant rootstocks used in the warm regions (i.e. Ramsey and 1103 Paulsen). Compared with the standards, must of fruit from the new rootstocks had lower pH, which has been linked to lower levels of K+ in the fruit (Clingeleffer 1996). Shiraz wines from the new rootstocks generally had enhanced quality attributes (i.e. higher levels of colour and phenolics despite being harvested at lower maturities (i.e. 23.7–24.5°Brix) than the standard rootstocks (25.2–25.6°Brix) over 3 years (2001–2003). Adoption of these rootstocks may facilitate integrated approaches to vine management involving high density plantings and low-input, lighter pruning systems to improve productivity (yield /ha) and water use efficiency, enhance wine quality attributes and reduce production costs, not only in the warmer regions but in cool climates where high input, vertical shoot positioned systems are used.
To remain economically sustainable, the Australian winegrape industry must address the challenges of consistent supply and quality in a variable and changing environment, continue to increase production efficiency and meet requirements to reduce alcohol content of Australian wines. The results indicate that a significant body of knowledge is available to implement an integrated, low-input total management systems approach for crop control and manipulation of fruit composition to modulate seasonal impacts, improve water use efficiency and reduce alcohol levels of Australian wines. Such a strategy will involve control of shoot vigour and the development of open canopies with development of small bunches with small berries which have with enhanced colour, phenolics and sensory appeal (Clingeleffer and Sommer 1995, Clingeleffer et al. 1999). These include lighter pruning to ensure productivity is maximised in seasons of low fruitfulness, deficit irrigation techniques and the adoption of low-to-moderate vigour rootstocks which also have potential to facilitate adoption of high density plantings to optimise yield and water use. Application of crop thinning techniques to both modulate yield levels and to promote early ripeness, colour and flavour development provides a further tool for integration into a total systems approach to vineyard management.
The results presented also identify major gaps in knowledge with regard to the environmental cues that are responsible for the large season to season variability in yield and composition. To meet the future challenges, particularly management of the seasonal variability in yield and fruit composition, further underpinning research is required to enhance our understanding of the causes of yield variability between seasons. This will involve identification of critical control points and identification of the critical external factors influencing crop development, in particular the main driver of yield (i.e. bunch number) and fruit composition. These studies must account for the differences between varieties and regions (i.e. the genotype × environment interaction). There is also a significant need to integrate current knowledge to identify the best approaches to manage the seasonal variability with the aim to have consistent supply and composition. Development of computer-based modelling software will be useful in this regard. The development of rapid techniques for post-set crop prediction to facilitate imposition of crop control treatments is also an urgent requirement.