Predicting Fruit Set in Australian Pinot Noir

It is extremely difficult, if not impossible, to predict grapevine yield from year to year, as there are many sources of variation from grape variety to climate that significantly increase the uncertain outcome of grape yield each growing season.  In Victoria, Australia, it has been shown that the greatest variation in grapevines occurring from season to season is bunch number per vine.  Knowing how many bunches of grapes per vine they will get every year is near impossible to predict using current methods.

What determines the number of bunches per vine?  Well, there are many factors involved that determine how many bunches of grapes per vine any one grapevine may have, including the number of flowers produced, the number of buds undergoing successful fruit set, as well as the potential number of bunches per bud.  In terms of how many buds will actually break to become potential fruit, a lot rides on the grape variety and the specific clone of grape.

Additionally, the specific wine region, such as New Zealand or France, is also important for potential fruit set, as the climate or weather conditions in that region will influence how my buds break, flower, and become fruit..  How well the vine grows after winter dormancy also plays a role in potential fruitfulness, which is influenced by how much energy the plant has in reserves during this quiescent time.

The short study presented today aimed to examine a statistical model for predicting fruitfulness in Pinot Noir grapevines, the results of which could be very important for the wine industry in terms of predicting yield at the beginning of the season as well as knowing what, if any, actions would be necessary to potentially increase this yield if need be.

Methods

Commercial vineyards in Southern Tasmania were used for this study: one in Derwent Valley, one in Coal River Valley, and one in Rokeby.  Two different clones of Pinot Noir were used: D5V12 and 114.  Grapevines were grown on their own roots and were arranged in a vertical shoot-positioned system.  Planting densities were the same for all three sites (1.5m x 2.3m), though the ages of each site were different (6, 9, and 10 years old).

Vines were pruned by hand and were pruned to two 10 bud canes and a 2 bud replacement spur.

Are pruning, canes were weighed and buds were harvested for carbohydrate analysis.

Three weeks after bud break, the numbers of flowers were recorded for each vineyard and for each Pinot Noir clone.

The statistical model used to predict fruitfulness in Pinot Noir was a Poisson-based generalized linear model (GLM) with a logarithmic link function.  Site, clone, and bud position were considered in the model, in addition to any and all interactions between the three.

Results

• There were significant differences found between the number of flowers present after bud break and the actual fruitfulness of these buds later on in the season.
• There was great variation between the numbers of flowers present on the grapevines between each site.
• Bud positions as well as site location were highly predictive of the number of flowers present on the vines after bud break.
• Bud position, site, and clone were all significantly predictive of the actual numbers of flowers, with significant interactions between site and bud position as well as site and clone.
• Pinot Noir clone D5V12 had lower probability of undeveloped buds, and also a decreased probability of multiple flowers per node.
• Cane weight and number of flowers per node were positively correlated (i.e. heavier canes = more flowers per node).
• Cane weight and percentage of stored starch were not correlated.
• There were no differences in stored starch levels between clones at the different sites.
• Starch levels were found to be a significant predictor of the number of flowers per node on the grapevines.
• For every 1 unit increase in starch, there was an associated bunch increase of 3.2% noted in Pinot Noir.
• The number of growing degree days was highest at the vineyard site that had the greatest fruitfulness.

Conclusions

The results of this short study indicate that there is significant variability between different vineyard sites in terms of fruitfulness in Pinot Noir.  While there was this strong variability between sites, it was determined that starch levels were highly predictive of number of flowers as well as number of grape bunches regardless of site or clone.  Those sites with higher levels of starch in the vines showed greater performance in terms of flowers after bud break as well as fruitfulness.

  Since starch is the primary source of carbohydrates for grapevines in the early spring (after winter dormancy), it seems as though those grapevines that make it through the winter with greater stores of starch will have the highest probability of successful flowering and higher fruit set.

Overall, though there was a lot of variability between site and clones, the results of this study indicate that it may be very important to control starch levels in grapevines prior to winter dormancy, in order to ensure great enough energy stores for the start of the following growing season.

This was a relatively small study, so before any sweeping conclusions can be made, the study should be repeated on a larger scale, in different grape growing regions, and on many more varieties.  The results of this study are promising in that starch and “carbohydrate” control during the season may be a good way of increasing one’s chances of good fruit set the following year, though of course more work needs to be done to confirm starch is a strong predictor of fruitfulness as well as what methods could be employed to increase starch stores for the winter.

I’d love to hear what you all think of this topic!  Please feel free to leave your comments and discuss!

Source: Jones, J.E., Lee, G., and Wilson, S.J. 2013. A statistical model to estimate bud fruitfulness in Pinot noir. American Journal of Enology and Viticulture 64(2): 274-279.