May 2016 Issue of Wines & Vines

Benefits and Costs of Early Leaf Removal

Researchers study ways to reduce bunch rot while maintaining yields

by Bryan Hed and Michela Centinari
Cluster Comparison
Cluster architecture plays a major role in bunch rot susceptibility. The small, overcrowded Vignoles cluster (left) is extremely susceptible, whereas the open, loose architecture of Chambourcin (right) generally renders it free of bunch rot in most years.

The development of bunch rot disease is very dependent on climate, especially when there is frequent rainfall and high humidity during the ripening period—a common scenario in Pennsylvania and other parts of the eastern United States. In addition, ripening generally coincides with the onset of hurricane season, which can often deliver abundant rainfall and cloud cover throughout the eastern seaboard region at that time. Unfortunately, there’s not much we can do about the weather, but there are ways to improve rot control in spite of that.

Fungicides are an important part of bunch rot control programs. The ubiquitous fungus Botrytis cinerea is often the predominant cause of late-season bunch rots, and fruit rot control programs rely heavily on Botrytis spp.-specific fungicide applications made at bloom, pre-bunch closure, véraison and pre-harvest (Wilcox 2012). However, pesticides are not always enough to maintain commercial levels of control, and Botrytis spp.-specific fungicides generally carry a high risk of the development of resistance, making heavy reliance on them unsustainable.

Furthermore, other organisms besides Botrytis are involved in the bunch rot complex (such as sour rot bacteria and other fungi) that are not well controlled by any pesticide options. Finally, growers are increasingly conscientious about public interest in reducing chemical inputs in agriculture. Therefore, bunch rot control strategy in the eastern United States must integrate non-chemical methods to be most consistently effective.


  • Researchers in Pennsylvania studied multiple ways to increase cluster looseness in order to reduce bunch rot.
  • Pre-bloom leaf removal was more consistent in reducing both the compactness of clusters and the potential for bunch rot.
  • While leaf removal at trace bloom can be done mechanically, it reduces the number of berries per cluster, and therefore reduces yield to some extent.
  • Yield reductions depend on the percentage of leaf area removed, the variety and weather conditions.

Conditions leading to bunch rot and possible solutions
Research has shown that the compactness of grape clusters plays a major role in bunch rot susceptibility (see photo at lower left). Indeed, the more compact a grape cluster, the more likely rot is to develop in that cluster (Vail and Marois 1991, Hed et al. 2009), and once initiated, the more likely it is to spread quickly and severely throughout the cluster (Zitter and Wilcox 2004). Berry cuticle, known as an important barrier to pathogens, was found to be thicker among berries grown in looser clusters with less berry-to-berry contact (Percival et al. 1993). Bloom “trash,” a mix of spent caps, flowers and stamens that can act as a substrate for bunch rot pathogens after bloom (Northover 1987), is less likely to be retained inside loose clusters, and when retained, is less likely to exacerbate bunch rot disease in loose clusters than compact clusters (Hed et al. 2009). Loose clusters may also improve pesticide spray penetration onto interior surfaces of clusters (Hed et al. 2011), which may improve pesticide efficacy—not only for control of Botrytis, but other diseases as well.

The development of bunch rots is strongly influenced by the fruit-zone microclimate, which can be modified to reduce fruit susceptibility. For example, leaf removal around clusters improves the microclimate in the fruit zone by increasing air circulation and cluster sunlight exposure, thereby reducing fruit wetness periods (English et al. 1989), and improving pesticide penetration into the fruit zone (see photos on page 85). The currently recommended timing of this widely adopted practice in cool-climate regions is generally from after fruit set to “pea-sized berries” stage in order to minimize potential reduction in yield and bud development at defoliated nodes (Candolfi-Vasconcelos and Koblet 1990, Wolf et al. 1986).

Post-fruit set leaf removal does not address the pitfalls of compact cluster architecture. However, when fruit zone leaf removal is applied earlier in the season—just before or at the beginning of bloom (pre-bloom; trace bloom)—most photosynthetically exporting leaves are removed, starving inflorescences of carbohydrates during bloom and reducing the number of flowers that set fruit (Coombe 1959, May et al. 1969, Candolfi-Vasconcelos and Koblet 1990). This in turn reduces the number of berries per cluster, cluster compactness and predisposition of clusters to bunch rots (Poni et al. 2006, Sabbatini and Howell 2010), combining the benefits of an open fruit zone with less disease-susceptible clusters (Hed et al. 2015).

In the past 10 years, pre-bloom leaf removal has gained popularity in some European countries as a practice for reducing crop level, improving grape composition and loosening clusters of high-yielding varieties such as Barbera, Sangiovese, Trebbiano and Tempranillo (Poni et al. 2006, Intrieri et al. 2008, Tardaguila et al. 2010). Recent studies conducted in Italy and Spain suggested that pre-bloom leaf removal could be a viable alternative to traditional cluster thinning for reducing crop level and improving grape and wine composition (Gatti et al. 2012, Tardaguila et al. 2012). Crop-regulation achieved with cluster thinning and pre-bloom leaf removal is based on two different mechanisms: The first reduces the number of clusters, the second cluster weight.

The increased interest in pre-bloom leaf removal over cluster thinning is related to its suitability for mechanization (Intrieri et al. 2008) and other desirable features such as improvement of canopy microclimate and decrease of cluster compactness. However, relying on pre-bloom leaf removal to consistently reduce yield to the desired level could be risky. In this sense, cluster thinning is a more conservative approach because, unlike pre-bloom leaf removal, its severity can be decided upon estimation of final yield. Furthermore, performing pre-bloom leaf removal in years when percent of fruit set is naturally reduced by unfavorable weather conditions could result in excessive crop reduction.

Bunch rot research at the LERGREC
Bunch rot research at the Lake Erie Regional Grape Research and Extension Center (LERGREC) in North East, Pa., initially utilized Vitis interspecific hybrid Vignoles, a highly susceptible white hybrid that produces small, compact clusters. In fungicide trials, rot control in this variety was often mediocre at best, even after the application of full season programs of highly active Botrytis spp.-specific fungicides such as Vangard, Switch and Rovral. After véraison, as Vignoles berries began to soften in overcrowded clusters, fruit on the insides of clusters would become crushed, and fruit on the outside would split or be partially pushed off of their stems. Damaged fruit rotted quickly, spreading rot extensively within the most compact clusters.

In 2001, the LERGREC began research to determine the role that cluster compactness played in bunch rot development in Vignoles, and to what extent the alleviation of compactness would reduce rot development in this variety. The results indicated strongly that the adoption of cluster loosening methods would increase the effectiveness of bunch rot control programs on this cultivar and possibly others (Hed et al. 2009).

Over the next several years, researchers examined a number of methods for their efficacy at reducing cluster compactness and bunch rot, with varying levels of success. The application of mineral- and vegetable-based oils and anti-transpirants just before bloom—in an attempt to mimic the effects of pre-bloom leaf removal by temporarily impairing photosynthesis during bloom and the early fruit-set period—met with limited success. Removal of the bottom third of each Vignoles cluster at pre-closure (clipping) was extremely effective at reducing rots, but also extremely unpractical. We also examined the effects of gibberellin applications, either pre-bloom (aimed at lengthening clusters to reduce compactness) or during bloom (to reduce compactness by reducing fruit set), the latter of which seemed to provide more consistent control of bunch rots than pre-bloom gibberellin (Hed et al. 2011) or oils and anti-transpirants. However, no method matched that of pre-bloom leaf removal in terms of consistent reduction in compactness and bunch rots.

Researchers at LERGREC initiated a long-term trial in 2007 with the two most successful of the cluster loosening treatments: fruit-zone leaf removal at trace bloom (when the first caps fall off at the beginning of bloom) and bloom applications of gibberellin (Hed et al. 2015). They compared these treatments with no leaf removal/no gibberellin (check) and with two other, later timings of leaf removal: just after fruit set (the currently recommended timing) and véraison.

Field trials were carried out using vigorous, six-year-old grapevines of clone 96 of Vitis vinifera Chardonnay grafted onto 3309C rootstock. Vine spacing was established at 1.8 meters within rows by 2.8 meters between rows. During dormant pruning, four canes were left: two canes at ~1 meter and two canes at 1.3 meters. Shoots were trained vertically between catch wires, and overcrowding of shoots was avoided by adjusting shoot number to between 16.4 and 19.7 shoots per meter row during early June. The experiment was set up in randomized complete blocks with four replications. No nitrogen fertilizer was applied, as vigor was more than adequate, and all experimental plots (12 vines each) received Botrytis spp.-specific fungicide applications at pre-closure and véraison.

Leaf removal research yields results
Researchers at LERGREC found that the earlier leaf removal was performed, the greater the reduction in Botrytis bunch rot, though the improvement was not always statistically significant (see table on page 84). For example, leaf removal at trace bloom, post-fruit set or véraison reduced Botrytis severity by an average of 71%, 47% or 14%, respectively, compared to no leaf removal. Leaf removal at trace bloom significantly reduced Botrytis incidence (percent clusters with rot) and severity (percent area of clusters with rot) in five and four seasons, respectively, and was as effective as two additional Botrytis fungicide applications (at bloom and at pre-harvest), suggesting potential to reduce fungicide inputs.

The results also showed that delaying leaf removal until véraison provided little to no benefit, with an average bunch rot-reduction of only 14% compared with no leaf removal. In harmony with our earlier studies, Botrytis development increased with the number of berries per cluster and berries per centimeter of cluster (compactness). The number of berries per cluster was significantly reduced by leaf removal at trace bloom in 2007, 2010 and 2011 (see photos on page 87), and by 10 mg/L gibberellin in 2007 and 25 mg/L in 2011. The number of berries per centimeter was significantly reduced by leaf removal at trace bloom and gibberellin in 2007, 2008 and 2011 (see table on page 83).

In general, bloom gibberellin sprays were less effective at reducing rots than leaf removal at trace bloom (see table on page 84). Like leaf removal at trace bloom, gibberellin sprays at bloom reduced harvest rots in almost every year, but reductions were significant only in 2008 at 5 mg/L and in 2010 at 25 mg/L, for incidence and severity, respectively. It should be noted that gibberellin is labeled for application to some seeded wine grape varieties during early pre-bloom stages, but the bloom application is experimental and not on the label.

Unfortunately, in the trials at LERGREC, applications made according to the label recommendations were not as effective as bloom applications. The label restrictions are designed to avoid the potential for negative effects on bud development that can occur on some varieties in the subsequent year after application. We experienced this in our studies with Riesling during 2010 and 2011, when bloom sprays of gibberellin at 25 mg/L resulted in excessive shot berries and reduced cluster development and yield the following season. Conversely, no negative effects were observed on return yield in our studies with Vignoles at similar rates (Hed et al. 2011).

These and other studies (Dass and Randhawa 1968, Pearson and Riegel 1983) confirmed that the effects of gibberellin application were complex and dependent on variety, rate, timing and even season. The low rates of gibberellin used in the Chardonnay study had inconsequential effects on yield, but there was variability in the beneficial effects on Botrytis bunch rot and cluster development from season to season. Nevertheless, when averaged over the six seasons, bloom gibberellin applications of 10 and 25 mg/L to Chardonnay reduced rot more and cost less than two additional Botrytis-specific fungicide applications; some incentive to define the spectrum of activity of gibberellin more clearly as well as its potential usefulness on other varieties (particularly hybrids), for which there is limited information.

None of the treatments had significant effects on juice composition, though average Brix trended higher numerically in leaf removal at trace bloom in comparison to the check in almost every year. The one exception was leaf removal at post-fruit set, which reduced titratable acidity in 2007, 2011 and 2012 when compared to the check.

Trace bloom leaf removal significantly reduced yield in 2008, but not in subsequent seasons (2009-12), and did not reduce return bloom (the number of clusters per shoot). These results may be partly an effect of leaving non-defoliated canes at the head of vines for renewal after 2007, as proposed by Sabbatini and Howell (2010). This practice minimizes the use of wood with buds whose potential yield has been reduced by early defoliation the previous season. There can also be some compensation in cluster weight reduction by an increase in berry weight. Trace bloom leaf removal in the six-year trial with Chardonnay resulted in an average increase in berry weight of about 6%.

Researchers were confident that potential negative year-after side effects could be managed, and similar trials with trace bloom leaf removal were conducted in commercial vineyards in southern Pennsylvania in 2010 and 2011 on Pinot Noir, Pinot Gris and Chardonnay. These trials showed similar results—reductions in cluster compactness and rot. However, the rot reductions in commercial vineyards were generally smaller, an effect attributed to: 1) the fact that all treatments received full season Botrytis fungicide programs (rather than the abbreviated program in the six-year Chardonnay trial above), and 2) that rot developed in some vineyards as a result of bird damage.

Next steps
Leaf removal in these studies has utilized hand labor. With the high cost and growing scarcity of labor, mechanization of trace bloom leaf removal is an important next step to improve cost effectiveness and adoptability of this practice (see photo on page 88). To date, the examination of air pulse leaf-removal technology using machines from Collard and BlueLine has met with some encouraging results: Researchers are observing many of the same beneficial effects as with trace bloom leaf removal by hand, though to a lesser degree. We are hoping further research will improve that outcome.

Air pulse leaf removal efficiency has ranged from 32% to 60% and was most efficient on vertical shoot position and four-arm-kniffen trellis systems, as opposed to more three-dimensional systems like high-wire no-tie (a very common, low-cost trellis system for hybrid grapes in Pennsylvania). Fortunately, the pulse air technology does not appear to damage inflorescences when applied just before bloom, and future assessments of return bloom will determine the potential for year-after effects of mechanized trace bloom leaf removal on bud development.

Early leaf removal and yield
Growers interested in applying pre-bloom/trace bloom leaf removal in their vineyards need to consider a number of factors before doing so. Trace bloom leaf removal intentionally reduces the number of berries per cluster (by an average of about 20% in our six-year study with Chardonnay) and can therefore reduce yield to some extent. The potential for negative effects on yield (Poni et al. 2006) may hinder their acceptance on varieties where crop reduction is undesirable or unnecessary.

Among other factors, yield reductions will depend on the percentage of leaf area removed, variety and weather conditions. For example, pulling just three to four leaves per shoot may do little to reduce cluster compactness on an extremely tight variety such as Pinot Gris, and pulling seven to eight leaves per shoot may reduce fruit set (and yield) too much. Therefore, it may be advisable to apply this treatment to just a few vines initially, alternating “treated” vines with “non-treated” vines (as checks for comparison) along, say, 10 or 20 vines in a row.

Cane pruning using non-defoliated renewals from the previous year will eliminate negative effects on bud development from early defoliation (Sabbatini and Howell, 2010). However, mechanization will preclude the use of non-defoliated renewals. Leaving more buds can mean more clusters and less reduction in yield. Managing these factors enabled researchers at the LERGREC eventually to eliminate yield reductions in their trial with Chardonnay while maintaining rot reductions at harvest, particularly in years when fruit-set (and cluster compactness) was naturally high.

As the wine grape industry in the eastern United States continues to expand, so do the challenges to the sustainability and consistency of bunch rot control strategies in susceptible wine grape varieties. Coupled with rising demand for fewer pesticide inputs and the never-ending challenges associated with managing pesticide resistance, the integration of cluster loosening methods may address some of these challenges while reducing reliance on pesticides in cool climate regions. The mechanization of these methods can provide interested growers with a more sustainable and cost effective, non-chemical approach to premium wine grape production, in spite of the region’s weather.

Bryan Hed is a research technologist at the Lake Erie Grape Research and Extension Center in North East, Pa. Dr. Michela Centinari is assistant professor of viticulture at The Penn State College of Agricultural Sciences. They would like to thank the Pennsylvania Wine Marketing and Research Board and the New York Wine and Grape Foundation for funding much of the bunch rot research at the Lake Erie Regional Grape Research and Extension Center.

Literature Cited
Candolfi-Vasconcelos, M.C., and W. Koblet. 1990. Yield, fruit quality, bud fertility and starch reserves of the wood as a function of leaf removal in Vitis vinifera – Evidence of compensation and stress recovering. Vitis 29:199-221.

Coombe, B.G. 1959. Fruit set and development in seeded grape varieties as affected by defoliation, topping, girdling, and other treatments. Am. J. Enol. Vitic. 10:85-100.

Dass, H.C., and G.S. Randhawa. 1968. Effect of gibberellin on ‘Golden Queen’ grape, a vinifera-labrusca hybrid. Am. J. Enol. Vitic. 19:52-55.

English, J.T., C.S. Thomas, J.J. Marois, and W.D. Gubler. 1989. Mi¬croclimates of grapevine canopies associated with leaf removal and control of Botrytis bunch rot. Phytopathology 79:395-401.

Gatti, M., F. Bernizzoni, S. Civardi, and S. Poni. 2012. Effects of cluster thinning and preflowering leaf removal on growth and grape composition in cv. Sangiovese. Am. J. Enol. Vitic. 63: 325-332.

Hed, B., H.K. Ngugi, and J.W. Travis. 2009. Relationship between cluster compactness and bunch rot in Vignoles grapes. Plant Dis. 93:1195-1201.

Hed, B., H.K. Ngugi, and J.W. Travis. 2011. Use of gibberellic acid for management of bunch rot on Chardonnay and Vignoles grape. Plant Dis. 95:269-278.

Hed, B., Ngugi, H. K., and Travis, J. W. 2015. Short- and Long-Term Effects of Leaf Removal and Gibberellin on Chardonnay Grapes in the Lake Erie Region of Pennsylvania. Amer. J. Enol. Vitic. 66: 22-29 

Intrieri, C., I. Filippetti, G. Allegro, M. Centinari, and S. Poni. 2008. Early defoliation (hand vs. mechanical) for improved crop control and grape composition in Sangiovese (Vitis vinifera L.). Aust. J. Grape Wine Res. 14:25-32.

May, P., N.J. Shaulis, and A.J. Antcliff. 1969. The effect of controlled defoliation in the Sultana vine. Am. J. Enol. Vitic. 20:237-250.

Northover, J. 1987. Infection sites and fungicidal prevention of Botrytis cinerea bunch rot of grapes in Ontario. Can. J. Plant Pathol. 9:129-136.

Pearson, R.C., and D.G Riegel. 1983. Control of Botrytis bunch rot of ripening grapes: Timing applications of the dicarboximide fun¬gicides. Am. J. Enol. Vitic. 34:167-172.

Percival, D.C., J.A. Sullivan, and K.H. Fisher. 1993. Effect of cluster exposure, berry contact and cultivar on cuticular membrane forma¬tion and occurrence of bunch rot (Botrytis cinerea PERS.:FR.) with three Vitis vinifera L. cultivars. Vitis 32:87-97.

Poni, S., L. Casalini, F. Bernizzoni, S. Civardi, and C. Intrieri. 2006. Effects of early defoliation on shoot photosynthesis, yield compo¬nents, and grape composition. Am. J. Enol. Vitic. 57:397-407.

Sabbatini, P., and G.S. Howell. 2010. Effects of early defoliation on yield, fruit composition, and harvest season cluster rot complex of grapevines. HortScience 45:1804-1808.

Tardaguila, J., F. Martiniez de Toda, S. Poni, and M.P. Diago. 2010. Impact of early leaf removal on yield and fruit and wine compostion of Vitis vinifera L. Graciano and Carignan. Am. J. Enol. Vitic. 61: 372-381.

Tardaguila, J., J.A. Blanco, S. Poni, and M.P. Diago. 2012. Mechanical yield regulation in wine grapes: comparison of early defoliation and crop thinning. Aust. J. Grape Wine Res. 18: 344–352.

Vail, M.E., and J.J. Marois. 1991. Grape cluster architecture and the sus¬ceptibility of berries to Botrytis cinerea. Phytopathology 81:188-191.

Wilcox, W.F. 2012. Vineyard disease management. In 2012 New York and Pennsylvania Pest Management Guidelines for Grapes. T.H. Weigle and A.J. Muza (eds.), pp. 11-38. Cornell and Penn State Cooperative Extension, New York.

Wolf, T.K., R.M. Pool, and L.R. Mattick. 1986. Responses of young Chardonnay grapevines to shoot tipping, ethephon, and basal leaf removal. Am. J. Enol. Vitic. 37:263-268.

Zitter, S.M., and W.F. Wilcox. 2004. Ontogenic, physical, and cultural factors affecting the initiation and spread of Botrytis bunch rot of grapes. In Proceedings of the Thirteenth Botrytis Symposium. N. Delen and Y. Elad (eds.), p. 28. Antalya, Turkey.

Print this page   PRINTER-FRIENDLY VERSION   »
E-mail this article   E-MAIL THIS ARTICLE   »
Currently no comments posted for this article.