For those of us who read the back labels of wine bottles we often hear that high quality wines can only come from vines with low yields, but is this always the case? There is enough experience, both practical and research-based, to show that the effects of yield are very much site-based and therefore largely dependent upon vine vigor or balance. The most straightforward metric of balance is crop load (Ravaz Index), which is the ratio of yield per vine vs. vine size (weight of cane prunings). A Ravaz Index between 5 and 12 is normally accepted as balance, whereas numbers below this are considered too vigorous, while values > 12 are normally considered overcropped. Experience has shown that varieties such as Pinot noir tend to be very sensitive to overcropping and therefore Ravaz Indices closer to 5 are accepted as ideal, whereas aromatic white varieties such as Riesling can normally be cropped at Ravaz Indices around 10-12.

There have been dozens of crop thinning studies done worldwide on many varieties. Generally the detrimental effects of high yield are: 1) Most pronounced in cool climates, because of delayed maturity; 2) Vintage dependent; 3) Site and variety dependent. When I worked in BC, we did crop control studies on Riesling, Gewurztraminer, and Pinot noir. Lowering crop level in Pinot noir had consistent positive effects, but cluster thinning Riesling and Gewurztraminer gave ambiguous results. Our research group has done work in Ontario on cluster thinning in Chardonnay musqué, Merlot, Cabernet franc, and Cabernet Sauvignon. Reducing crop load in Bordeaux red varieties in Niagara enhanced color, phenols, and wine quality in one cool season but was less effective in warmer seasons. With Chardonnay musqué-lowering crop enhanced terpenes (compounds responsible for muscat flavor) and increased wine quality in cool seasons but had minimal effects in warm years.

Chapman et al. (2004) in California studied the role of yield in the sensory properties of Cabernet Sauvignon using pruning and cluster thinning to manipulate yield. Cabernet Sauvignon vines were subjected to six winter pruning treatments over two years and eight cluster-thinning treatments over one year, with thinning imposed at veraison. These treatments created yields that varied from 4.3 to 22.2 t/ha. Wines made from vines pruned to low bud numbers (hence "low yield") were higher in "veggie" aroma and flavour, bell pepper aroma, bitterness, and astringency than "high-yield" wines. Conversely, the wines made from vines pruned to high bud numbers were higher in red/black berry, jam and fresh fruit aromas, and fruity flavour than low-yield wines. In general, "veggie" attributes decreased in intensity and fruity attributes increased in intensity as node number and yield increased.

A recent study by Nick Dokoozlian at Gallo Wines in California looked at the effect of crop load treatments (undercropping, balanced, overcropping) on the relationship in Cabernet Sauvignon between harvest Brix and concentrations of methoxypyrazines, the compounds responsible for the veggie aroma. In both overcropped and undercropped vines, the fruit was at nearly 29 oBrix before it had dropped below the methoxypyrazine aroma threshold when it could be harvested, but in balanced vines, the threshold was reached at 24 °Brix. The conclusion was that fruit oBrix and flavor are much more "in synch" in vines that are balanced.

Recently our group investigated the impacts of naturally-varying yields on wine composition and sensory quality of Riesling and Cabernet franc. No cluster thinning was done. The sites were located in five of the 10 VQA sub-appellations. Cultural practices, including pruning and canopy management were consistent across each vineyard block. Vine yields were divided into equal-sized categories of "low", "medium", and "high" at harvest (with three replicates at each site). Since sites were different in terms of vigor, ranges in yields were also different; e.g. high and low Riesling yields (category averages) ranged from 6.7 to 3.7 kg/vine at the highest-yielding site to 5.0 to 3.0 kg/vine at the lowest-yielding site Yield was largely influenced by number of clusters/vine, and was also associated with larger berries and higher Ravaz Indices. All resulting wines were analyzed for pH and titratable acidity (TA) and the Cabernet franc wines were also analyzed for color hue and intensity, anthocyanins, and total phenols. All replicate wines were successfully sorted by a sensory panel to confirm differences between wines produced from different yield categories and between different sites. All wines were thereafter subjected to descriptive analysis.

As a general rule we found that at some sites crop yields could reach as high as 7-8 tonnes/per acre before overcropping had an impact on wine quality. Low yield Cabernet franc wines had higher color, anthocyanins and total phenols but natural yield variance had no impact of yield on TA or pH in either variety. Increasing yield in Cabernet franc reduced concentration of sec-butyl methoxypyrazine and increased isobutylmethoxypyrazine (IBMP). IBMP increased with increased crop level in three vineyards (Beamsville Bench, Lincoln Lakeshore north, Lincoln Lakeshore south). Five aroma and eight flavor attributes for the Riesling and three aroma and six flavor attributes for the Cabernet franc differed amongst the different yield categories in the descriptive analysis. High yield Riesling wines at some sites were associated with lower intensity of fruit characteristics and higher mineral or floral attributes. Cabernet franc vines with higher yield, vine size, and clusters/vine often produced wines with higher bell pepper characteristics and less fruit characteristics.

Sub-appellations also had an effect on the wine sensory profile. The Lincoln Lakeshore (North) and Niagara Lakeshore Riesling vineyards were associated with higher mineral characteristics, the Four Mile Creek location had more apple/pear characteristics, and the St. Davids Bench, Beamsville Bench and Lincoln Lakeshore (South) vineyards were highest in fruit and citrus characteristics. Cabernet franc vineyards located on the Beamsville Bench and the Four Mile Creek sub-appellations were highest in bell pepper aroma, the Lincoln Lakeshore (north) vineyard was associated with more earthy character, and the Lincoln Lakeshore (south) sub-appellation was associated with more cooked fruit character.

What is the take-home message? Crop load effects differ from site to site, partly because of the variability in yield between sites, as well as the magnitude of variability at each site. Sites with low yield variability are less likely to display effects of yield. These data were based on one season-2010-which was quite warm. We plan to repeat this for the 2011 wines this upcoming spring.

Thanks to Lee Baker, who did all the sensory analysis,Mary Jasinski for geomatic analysis, and to Gary Pickering and Susanne Kogel for IBMP analysis