grapegrowing
The need for a risk-based approach to botrytis management Jacky Edwards
Robert Beresford
Biosciences Research Division, Department of Primary Industries 621 Burwood Highway, Knoxfield, Victoria 3180
The New Zealand Institute for Plant and Food Research Limited (Plant & Food Research) Mt Albert Research Centre, Private Bag 92-169 Auckland, NZ
Gareth Hill David Riches
The New Zealand Institute for Plant and Food Research Limited (Plant & Food Research)
Biosciences Research Division, Department of Primary Industries
Peter Wood
Kathy Evans
Plant & Food Research, Hawke’s Bay Research Centre Private Bag 1401, Havelock North, Hastings 4157, NZ
Tasmanian Institute of Agricultural Research, University of Tasmania New Town Research Laboratories, 13 St Johns Avenue, New Town, Tasmania 7008.
Dion Mundy Plant & Food Research, Marlborough Wine Research Centre PO Box 845, Blenheim 7240, NZ
The take-home message of this article is that vineyard managers should not automatically assume that fungicide treatments are the most effective and profitable way of managing botrytis risk in cool climate viticulture. Sometimes disease pressure is too low to warrant the costs of any treatment. Twenty-three field trials were conducted during 2006-08 across six cool climate regions of Australia and New Zealand; only eight of them developed more than 3% botrytis severity at harvest in the untreated plots. The potential for economic losses from botrytis in a given vineyard is determined partly by regional climate and partly by seasonal weather. The best disease management strategy for a particular region is determined by the historical likelihood of botrytis, by the risk imposed by viticultural practices (e.g. the level of bunch exposure) and by the risk of fungicide residues. The tactical decisions about the need for individual sprays or canopy management actions are determined by the way that a season develops. In the future those tactical decisions will be supported by a validated model that summarises all the information relevant to botrytis risk in the current season. A prototype model will be available soon for further development as a user-friendly tool. In the meantime, canopy manipulations such as leaf plucking, which do not use water or disrupt IPM programs used for other pests and diseases, may prove as effective as fungicides. Introduction
In a bad year, botrytis bunch rot (referred to as botrytis in the rest of this article) reduces both grape yield and wine quality, resulting in serious economic losses. For example, the 2008 vintage in the Hunter Valley was very badly affected. Less than 10% of the red crop was harvested and the white crop was picked at lower Baumes than normal to minimise losses (Riley 2008). As an insurance policy, spray programs for botrytis control generally are designed for high disease pressure situations, but how effective are they really, and what about lower risk seasons or sites? In the current climate of rising fuel and pesticide prices, and low availability of water and labour, is there a better way to manage the botrytis risk? To address this, the grape and wine industries on both sides of the Tasman have joined forces. A collaborative research project is being conducted by the Tasmanian Institute of Agricultural Research, the Victorian Department of Primary Industries and Plant & Food Research, New Zealand (formerly HortResearch) on behalf of the Grape and Wine Research and Development Corporation and New Zealand Winegrowers to develop a prototype model for predicting the in-season risk of botrytis in cool climates. Through this trans-Tasman collaboration and previous research in NZ, a large dataset comprising of 44 site-years has been collected. Trials were situated in the following regions: Yarra Valley (Victoria), Coal River and Derwent Valleys (Tasmania) and Auckland, Marlborough and
6 The Australian & New Zealand Grapegrower & Winemaker
Hawke’s Bay (NZ). Data collected includes site location, variety, and standard measures of yield, weather information (temperature, relative humidity, surface wetness, rainfall), canopy density, bunch exposure, amount of latent Botrytis cinerea infection at pre-bunch closure, and botrytis development post-veraison. This information is being used to identify early and late season predictors of botrytis risk for development of a prototype model to predict risk (Beresford and Hill 2008). The ultimate goal is to develop a decision support system for vineyard managers that will enable them to tailor their disease management practices to suit the level of botrytis risk for their site and the season. To develop the dataset for model development, trials were designed using untreated plots and combinations of treatments to generate a wide range of harvest botrytis severities (from 0 - 40%). This generated a considerable amount of valuable information on the effectiveness of disease control measures during the course of the project. In particular, fungicide timing, canopy management (leaf plucking and shoot thinning) and inoculum reduction measures (bunch trash removal) were tested and the results are the focus of this article. Botrytis disease cycle
There are several pathways by which B. cinerea can initiate bunch rot and the relative importance of these is still a subject of scientific debate (reviewed in Evans 2008). The fungus can infect developing berries as early as flowering, but then becomes arrested by anti-fungal compounds in the young berry. This is known as latent infection. The fungus also colonises dying tissues, so can be found in bunch trash such as caps, aborted berries and other debris that collects within a bunch after flowering. Once grape berries reach veraison, the anti-fungal compounds disappear and if environmental conditions are favourable, latent infections resume growth and B. cinerea in colonised bunch trash is able to infect berries. Both the
Fig. 1. Floral bunch trash colonised by Botrytis cinerea following moist incubation (left) and bunch trash removal in a trial plot using compressed air (right).
www.winebiz.com.au
Annual Technical Issue 2009
grapegrowing Table 1. Effect of removing bunch trash on botrytis severity at harvest for Australian and New Zealand field trials. Botrytis Severity at Harvest (%) Year
Region
Variety
Untreated control
Bunch trash removal
2006-2007
Auckland (NZ)
Sauvignon Blanc
25.3
25.1
2006-2007
Marlborough (NZ)
Sauvignon Blanc
0.2
0.1
2006-2007
Southern Tasmania
Sauvignon Blanc
2.5
2.5
2007-2008
Southern Tasmania
Chardonnay
3.6
4.8
2007-2008
Yarra Valley (Vic)
Chardonnay
2.0
2.7
latent infection and bunch trash pathways lead to the same endpoint – botrytis bunch rot. Post-veraison, B. cinerea can spread from berry-to-berry, and of course there is also the possibility of new infections through any damage on the ripening berries. Disease management treatments are designed to target these pathways. Fungicide applications at flowering, pea size and prebunch closure protect developing berries; late season sprays protect the susceptible berries from new infections. Bunch trash removal, a purely experimental treatment, was aimed at reducing inoculum within the developing bunch. Vine management techniques such as leaf plucking and shoot thinning modify the bunch zone microclimate to be less favourable for disease development and expression. Can removing bunch trash reduce botrytis risk?
Nine site-years of data from New Zealand showed a correlation between the amount of bunch trash colonised with B. cinerea and disease severity at harvest (Beresford and Hill 2008). It therefore seemed logical if the inoculum can be reduced by removing the bunch trash that disease risk will be reduced. However, when bunch trash was removed in our field trials (Figure 1), no significant reductions in botrytis severity occurred (Table 1). In fact, in some
trials there was increased disease in the trash removal treatments. We think this might be due to microscopic damage caused to the berries by the compressed air stream used to blow floral trash out of bunches, resulting in increased infection. Therefore, alternative methods of reducing inoculum in bunch trash such as colonisation by biological control agents may prove more effective. The method for quantifying the amount of infested bunch trash could be used to see if a treatment reduces botrytis inoculum and ultimately severity at harvest. How much botrytis control do fungicides give?
Fungicides have traditionally been the main component of botrytis control programs. In our field trials, fungicide applications were made at the following growth stages: 5% cap-fall, 80% cap-fall, pea size, pre-bunch closure (PBC), veraison and pre-harvest. The efficacy of early season (5%, 80% cap-fall), mid season (pea size, PBC), late season (veraison, pre-harvest) and combination treatments were compared with each other and with untreated plots. The results to date varied considerably across the different trial sites and seasons (Table 2). Many wineries impose a price penalty for botrytis-affected grapes MEA51167/4C/G&W
Water logged? You bet.
For more than 20 years, MEA has been designing simple, reliable and affordable solutions for ‘logging’ moisture levels in your soil. And the GDot is about as simple as it gets. Its bright yellow dots let you monitor your moisture levels from up to 15 metres away in most light conditions (you don’t even need to stop as you drive past!) Then there’s our GBug, a clever little fella
Annual Technical Issue 2009
that takes moisture readings every two hours. To access the data on a small hand-held device, simply wake him up with a ick of a nger. No fuss, no worries and no big bills to pay. Just simple solutions to help you save water. Talk to MEA directly or enquire about your nearest distributor. Call 08 8332 9044, email
[email protected] or visit www.mea.com.au.
www.winebiz.com.au
The Australian & New Zealand Grapegrower & Winemaker 7
grapegrowing if harvest severity is 3% or greater (Viti-Notes 2005). In 2006-07, only three out of ten sites exceeded 3% botrytis severity at harvest on unsprayed vines (Table 2). In two of these trials, full fungicide programs (4-6 sprays) were applied, but botrytis severity was only reduced below 3% at one of them. At the third site, no full spray treatment was applied. The early, mid and late season treatments (two sprays each) significantly reduced disease but were still above 3% at harvest. In 2007-08, five out of 13 sites exceeded 3% botrytis severity at harvest on unsprayed vines, but fungicide applications brought four of these sites to below 3% disease severity. Investigation of fungicide timing showed that there was no single growth stage where fungicide application consistently performed better. In some trials, 80% cap-fall applications gave the best disease control while in other trials the mid season applications (pea size to pre-bunch closure) performed better. Fungicides applied at veraison or later were not more effective than earlier applications.
Fig. 2. Leaf plucking significantly (P=0.033) reduced mean botrytis severity at harvest from 4.3% to 1.7%. Data from 10 trial sites in Victoria, Tasmania and New Zealand and two vintages (2007-2008) were combined.
Canopy management for botrytis control
Canopy density measurements, using the point quadrat method (Smart and Robinson 1991), were taken at some sites and showed leaf plucking was most effective at the sites with denser canopies i.e. dense canopies benefited more from leaf removal than did sparse canopies. Where leaf removal was combined with fungicide applications, these treatments were usually the most effective at reducing botrytis at harvest (Table 2). This effect appeared to be due to both the decreased suitability of the canopy microclimate for botrytis and improved fungicide penetration.
Leaf plucking from the bunch zone and shoot thinning are several of many management techniques that can open up the canopy to sunlight and wind, reducing the risk of botrytis development (Gubler et al. 1987). In 2006-07, leaf plucking was trialled at the New Zealand sites (Table 2) and shoot thinning at the Victorian sites. Due to the lack of disease development in the Victorian trials, no conclusions can be drawn about the effectiveness of shoot thinning. However, leaf plucking gave excellent results in the New Zealand trials, equal to or better than the fungicide treatments, so we decided to trial it in Tasmania and at all Victorian sites in 2007-08 instead of shoot thinning. In New Zealand, leaves were removed from both sides of the canopy; however in Australia leaves were only removed from the least exposed side of the canopy due to concern about sunburn, plus the desire to maintain acidity in Riesling grapes grown in Tasmania. Once again, the New Zealand sites gave excellent results, as did one site of Sauvignon Blanc in Victoria. In the ten trials that included leaf plucking treatments, the mean harvest botrytis severity for leaf plucked plots (1.7%) was significantly less than the mean for the non-plucked plots (4.3%) (Figure 2).
Trial results in relation to prototype prediction model
Data analyses for the modelling component of our project showed that botrytis severity at harvest is strongly correlated with the length of the ripening period (Beresford and Hill 2008), and that 3% severity is reached, on average, at approximately 40 days postveraison (Figure 3). Auckland and Hawke’s Bay were most prone to serious botrytis epidemics, followed by Marlborough and Southern Tasmania, with Victoria the lowest risk, where disease severity at
Table 2. Botrytis severity at harvest in untreated, leaf plucked and the most effective fungicide treatment in cool climate botrytis trials (2006-2008). At the time of writing, harvest data for 2008-2009 were being collected. A dash (-) indicates that the treatment was not tested at the trial site. Mean botrytis severity at harvest (%) Year
Region
Variety
Untreated
Full fungicide program*
Leaf plucked
2006-2007
Auckland (NZ)
Sauvignon Blanc
25.3
7.7 (mid)
-
Fungicide+leaf plucking -
“
Hawke’s Bay (NZ)
Sauvignon Blanc
4.5
2.1
2.6
0.3
“
Hawke’s Bay (NZ)
Chardonnay
9.6
5.9 ns
2.2
1.7
“
Marlborough (NZ)
Sauvignon Blanc #1
0.2
0.1 ns (early)
-
-
“
Marlborough (NZ)
Sauvignon Blanc #2
0.6
-
-
-
“
Southern Tasmania
Sauvignon Blanc
2.5
0.9 (mid)
-
-
“
Southern Tasmania
Riesling
0.6
0.8 ns
-
“
Yarra Valley (Vic)
Chardonnay #1
0.0
0.0 ns
-
-
“
Yarra Valley (Vic)
Sauvignon Blanc
0.1
0.0 ns
-
-
“
Melbourne (Vic)
Chardonnay
0.0
0.0 ns
-
-
2007-2008
Auckland (NZ)
Sauvignon Blanc
1.6
0.1 (early)
0.7
0.3
“
Hawke’s Bay (NZ)
Sauvignon Blanc
8.4
3.5
3.4
0.4
“
Hawke’s Bay (NZ)
Chardonnay
9.5
1.9
1.9
0.3
“
Marlborough (NZ)
Sauvignon Blanc #1
2.1
1.5 (mid/early)
1.1
0.2
“
Marlborough (NZ)
Sauvignon Blanc #2
0.8
-
-
-
“
Marlborough (NZ)
Sauvignon Blanc #3
5.3
0.6
-
-
“
Southern Tasmania
Sauvignon Blanc
2.2
1.6 ns (mid)
2.7
1.9
“
Southern Tasmania
Riesling #1
6.2
2.6 (mid)
-
-
“
Southern Tasmania
Riesling #2
1.9
0.8 ns (mid)
-
-
“
Southern Tasmania
Chardonnay
3.6
2.8 (late)
-
-
“
Yarra Valley (Vic)
Chardonnay #1
2.0
0.0
2.1
0.0
“
Yarra Valley (Vic)
Sauvignon Blanc
2.3
0.1
0.4
0.6
“
Yarra Valley (Vic)
Chardonnay #2
0.8
0.0
0.6
0.0
* Where a full fungicide program was not used in the trial, the most effective fungicide timing was chosen. The fungicide timing is shown in brackets. ns- fungicide treatments not significantly different to untreated at P ≤ 0.05 8 The Australian & New Zealand Grapegrower & Winemaker
www.winebiz.com.au
Annual Technical Issue 2009
grapegrowing
Fig. 3. Regional means of late-season interval (the ripening period from veraison to harvest) versus mean harvest botrytis severity in grapes. Regions are Victoria (V), Tasmania (T), Marlborough (M), Hawke’s Bay (HB) and Auckland (A). (Modified from Beresford and Hill 2008).v
harvest never exceeded 3% on untreated vines over the two seasons of trials. There was a broad range of weather and climatic conditions across the trial sites and seasons reported here. The results for 2008-09 are expected to again reflect this variation with Victoria experiencing its driest start to a year on record in 2009. Examination of rainfall patterns across the sites and seasons of our trials showed that late season rainfall (Figure 4) and mean daily rainfall (Beresford and Hill 2008) were significantly correlated with botrytis severity at harvest, but they had poor predictive ability. The duration of surface wetness, when used in a risk index that takes into account temperature during the wetness period (Bacchus index), has proved the most reliable weather factor for modelling botrytis risk (Beresford and Hill 2008). The botrytis prediction model currently under development aims to identify seasons with elevated botrytis risk as early as possible to allow control measures to be tailored to suit the level of risk and to prevent large scale crop losses such as those that occurred in the Hunter Valley, NSW, in 2008. Conclusions and future work
Assuming that 3% botrytis severity is a threshold level at which growers may suffer a price penalty on their crop, many of our trial sites did not reach the 3% severity threshold even when left untreated. In the eight trials where botrytis in the untreated plots exceeded 3% severity, fungicide applications brought the level down below the 3% threshold in only four of them. These results highlight the need for benefit-cost analyses, as the full cost of a spray program (fuel, pesticide, labour, water, etc) may not always be recouped, particularly in drier viticultural regions like Victoria. In wetter regions such as Hawke’s Bay, however, a fungicide program is likely to be the best strategy in most cases. Leaf removal from the bunch zone was often as effective as a complete fungicide program, particularly at sites in New Zealand, and this agrees with earlier New Zealand trial results (Agnew et al. 2004). Greatest benefits from leaf removal could be expected where the canopy is reasonably dense before leaf removal. In both Australia and New Zealand, the combined effect of leaf removal and fungicide application generally gave the lowest levels of botrytis. Trial results from 2008-09 are currently being collated and analysed for refining the prototype model. Once this has been completed, the next stage of development of the decision support system for botrytis management is testing and modification for commercial use, and identifying a suitable delivery platform for industry. In New Zealand, many grapegrowers (and other New Zealand farmers) have access to a network of weather-stations in major agricultural regions maintained by a commercial service provider, who provide access to weather information and disease
Annual Technical Issue 2009
Fig. 4. Late season rainfall and botrytis severity at harvest (logit transformed) across 23 site-seasons (2007-2008 vintages) in New Zealand and Australian cool climate vineyards. Late season rainfall is a significant contributing factor to Botrytis severity at harvest (P = 0.008), accounting for nearly 30% of the variance in logit harvest severity.
models via a subscription basis. Equivalent weather networks/ services will need to be developed in cool-climate grape growing regions in Australia. Acknowledgements
We wish to thank the Grape and Wine Research and Development Corporation, New Zealand Winegrowers, the Victorian Department of Primary Industries, the Tasmanian Institute of Agricultural Research, and Plant and Food Research, New Zealand, for funding this research. We also wish to thank the grapegrowers who made their vineyards available and assisted with the research trials used in this study. Special thanks to Bob Emmett for critical review of the article and to staff and students who assisted with trial work, including Rob Agnew, Michela Cambiotti, Justin Direen, Katie Dunne, Warwick Henshall, Kwang Soo Kim, Victoria Raw, Tine Tach, Tracy Taylor and Peter Wright. References Agnew, Mundy and Balasubramaniam (2004) Effects of spraying strategies based on monitored disease risk on grape diseases control and fungicide usage in Marlborough. New Zealand Plant Protection 57: 30-36. www.nzpps.org/ journal/57/nzpp57_030.php Beresford and Hill (2008) Predicting in-season risk of botrytis bunch rot in Australian and New Zealand vineyards. ‘Breaking the mould: a pest and disease update’, Australian Society of Viticulture and Oenology Seminar Proceedings, Mildura 24 July 2008. Evans (2008) Overview of R&D for managing botrytis bunch rot in Australia. ‘Breaking the mould: a pest and disease update’, Australian Society of Viticulture and Oenology Seminar Proceedings, Mildura 24 July 2008. Gubler, Marois, Bledsoe and Bettiga (1987) Control of botrytis bunch rot of grape with canopy management. Plant Disease 71: 599-601. Riley (2008) Practical management of bunch rots in high risk environments. ‘Breaking the mould: a pest and disease update’, Australian Society of Viticulture and Oenology Seminar Proceedings, Mildura 24 July 2008. Smart and Robinson (1991) Sunlight into wine, a handbook for winegrape canopy management. Winetitles, Adelaide, Australia; pp 88. Viti-Notes (2005) What wineries want ….and why: Winegrape assessment in the vineyard and at the winery. Grape purity 1. Diseases – powdery mildew, downy mildew, Botrytis and other moulds and rots. www.crcv.com.au/viticare/vitinotes
Contacts for further information: Dr Jacqueline Edwards, DPI Victoria:
[email protected] Dr Katherine Evans, University of Tasmania:
[email protected] Dr Robert Beresford, Plant & Food Research, NZ:
[email protected]
www.winebiz.com.au
The Australian & New Zealand Grapegrower & Winemaker 9
