Assessment of UV-C Radiation as an Integrative Pest Management Tool for the Management of Grape Powdery Mildew and Botrytis Bunch Rot

Progress report for GW21-219

Project Type: Graduate Student
Funds awarded in 2021: $30,000.00
Projected End Date: 07/31/2023
Host Institution Award ID: G224-22-W8615
Grant Recipient: Oregon State University
Region: Western
State: Oregon
Graduate Student:
Major Professor:
Expand All

Project Information


Management of grape powdery mildew requires great financial, environmental, and public health costs due to a heavy reliance on fungicides (~89% of the fungicide active ingredient used in grape production). New pest management tools are needed to reduce the environmental effects of heavy fungicide use and mitigate the emergence and spread of fungicide resistance. Germicidal ultraviolet C light (UV-C) may provide a new management tool for the improved control of powdery mildew in grapes that addresses these goals. Our research has shown that powdery mildew isolates vary in their tolerance to UV-C, but the effective doses are within ranges that can be reached in field applications. Small plot research trials demonstrated that UV-C significantly reduces disease development with or without augmentation of fungicides and without impacting fruit quality. Although treatments of UV-C alone exceeded commercially acceptable disease levels, it is likely this is an artifact of plot-to-plot interference from the untreated controls. It is possible that the UV-C applications will be more effective in a commercial setting where disease pressure is lower and more realistic. This technology needs to be tested at a commercial scale and compared to current management practices. However, this is not likely to occur until autonomous application technology (e.g., self-driving devices) reaches the robustness required for long-term field use. Finally, research results and live demonstrations of the UV-C application arrays have been made in commercial vineyards at various grower and researcher meetings serving Oregon, Washington, and California. Virtual and in-person research presentations and demos were well attended and accounted for both growers, researchers, and other agricultural professionals. Survey results show a high level of knowledge gained at events by growers who participated in the survey, and at least eight growers immediately indicated that there is interest in incorporating the technology into their disease management practices. Ten others have reached out for design parameters and two have visited to examine our research unit array.

Project Objectives:
  1. Identify the UV-C radiation dose and timing required to inhibit the growth of Erysiphe necator (grape powdery mildew) and Botrytis cinerea (bunch rot).
  2. Determine the suitability of UV-C radiation applications, under field conditions, for the management of powdery mildew and Botrytis bunch rot disease and its impact on fungicide-resistant pathogen populations.
  3. Examine the suitability of an autonomous robotic platform to deliver UV-C applications at a commercial scale.
  4. Determine the impact of UV-C application on fruit chemistry
  5. Education and demonstration of the tractor-mounted and robotic UV-C arrays directly to producers
[caption id="attachment_750971" align="alignright" width="300"]Gantt chart for the proposed completion of each objective. Q1: August-October; Q2: November-January; Q3: February-April; Q4: May-July. Shaded boxes indicate activities per quarter.[/caption]

Efficacy trials under laboratory conditions are already underway and will continue through into year two. The field studies at the BPP research vineyard and WVV will start in April and end in September or October after which point the tests on the harvested fruit chemistry will begin and then end before the beginning of the next field season. Demonstrations will be held in spring and summer to show the arrays in an accurate field setting.  Display of the UV-C applications units will also occur starting in Spring and continuing throughout the summer.


Click linked name(s) to expand/collapse or show everyone's info
  • Dr. Walter Mahaffee (Researcher)
  • David Markel - Technical Advisor - Producer
  • Dr. Dan Sargent - Technical Advisor (Researcher)
  • Alexander Wong (Researcher)


Materials and methods:

To examine the utility of UV-C radiation for the management of grape powdery mildew and bunch rot, investigated the following objectives:

  1. Identify the UV-C radiation dose and timing required to inhibit the growth of Erysiphe necator (grape powdery mildew) and Botrytis cinerea (bunch rot).
  2. Determine the suitability of UV-C radiation applications, under field conditions, for the management of powdery mildew and Botrytis bunch rot disease and its impact on fungicide-resistant pathogen populations.
  3. Examine the suitability of an autonomous robotic platform to deliver UV-C applications at a commercial scale.
  4. Determine the impact of UV-C application on fruit chemistry
  5. Education and demonstration of the tractor-mounted and robotic UV-C arrays directly to producers

Objective 1. Identify the UV-C  dose and timing required to effectively manage grape powdery mildew and Botrytis bunch rot

Effect of UV-C dose on conidia germination. Single conidium Erysiphe necator isolates were grown on detached grape leaves until sporulating (~10-14 days old) in a growth chamber at 21°C with a 16-hour photoperiod. Conidia deposited to Gelzan gum in a 24-well plate by puffing air onto the colony in a settling tower and allowing conidia to settle for 5 minutes.

Single conidium Botrytis spp. isolates were grown on potato dextrose agar amended with streptomycin (SPDA) for 10 days at 20°C under 16 hr. photoperiod of blacklight and fluorescent light to promote sporulation. Conidia were harvested by adding 3 mL of 0.05% Tween 20 and gently rubbing the culture with a sterile loop. Conidia were filtered through sterile cheesecloth and adjusted to 104 conidia per mL and 50 µL spread 0.5% Gelzan gellan gum medium based on Janisiewicz (2016). 

Conidia were incubated in the dark for 1 hour before being exposed to doses of UV-C radiation of approximately 0, 100, 150, 275, 500, and 800 Joules per meter squared (J/m2). To mimic how UV-C radiation was supplied in the field, two 16 in. 36-watt UV-C compact fluorescent lamps (GPL36K, Ushio) inside a retrofitted greenhouse lamp housing lined with aluminum foil and fitted with curtains of reflective foil placed over a conveyer belt. This system allows the dosage to be controlled by both the distance from the lamps to the sample and the speed of the sample passing under the lamps. To control which wells of the plate were exposed and protect the others, custom lids were 3D printed with gaps exposing only the desired wells. After exposure to UV-C radiation, the plates were incubated at 21°C for a darkness period of 7 hours, returning to the 16-hour photoperiod for 48 hours before being imaged with an Echo Revolve digital microscope (Figure 1). The images were processed in ImageJ software using the “HyphaTracker” macro to estimate the amount of hyphal area of each image (Brunk et al. 2018). Each dose of UV-C was applied to four wells on each plate, growth was averaged across the wells and dose-response curves were created in R using the ‘drc’ package to estimate the half-maximal effective dose (ED50) of UV-C to inhibit germination of powdery mildew and Botrytis conidia compared among isolates with different fungicide resistance profiles.

Microscope images of spores germinating, weakly germinating, or not germinating
Figure 1. Images of conidia 48 hours after exposure to UV-C with image backgrounds removed. A) Control without exposure to UV-C with strong germination growth. B) Exposure to 150 J/m^2 with weak germination growth, and C) exposure to 800 J/m^2 with no germination.

Botrytis growth was done in vitro by inoculating each well of a Gelzan gum media in 24-well plates with a 1 µL drop of 10,000 conidia per milliliter in a 0.05% Tween 20 solution. Plates were incubated at 20°C and 16-hour photoperiod, and the plates were exposed to UV-C at the above-described doses after an hour of darkness at the time of inoculation and 24 hours later. The wells were stained with florescent calcofluor to stain fungal chitin. The plate was fluorescently imaged 24 and 48 hours after UV-C exposure and the mycelial network was mapped using the software “Fungal Feature Tracker” to determine colony growth (Figure 2). Each dose of UV-C was applied to four wells on each plate, growth was averaged across the wells and dose-response curves were created in R using the ‘drc’ package to estimate the half-maximal effective dose (ED50) of UV-C to inhibit growth.

Workflow showing the raw image of stained Botrytis and the final mapped mycelium
Figure 2. Example of the Fungal Feature Tracker software “mapping” the Botrytis mycelial network. Top) the binary image, and Bottom) the image after being processed. Green indicates the hyphae, and pink is the hyphal tips. Growth is estimated as the length of the hyphal network.

Objective 2. Determine the suitability of UV-C radiation applications, under field conditions, for the management of powdery mildew and Botrytis bunch rot disease and its impact on fungicide-resistant pathogen populations.

Disease and resistance management of UV-C application. The small plot experiment using the tractor-mounted UV-C array was performed at the Botany and Plant Pathology research vineyard in Corvallis, OR. The vineyard consists of vertical shoot positioning trained Pinot Noir vines. The experimental design was a split-plot design with three replications in a randomized complete block. The UV-C array consists of 24, 55-Watt, 3 ft. fluorescent UV-C lamps (Sylvania Osram G55T8) with entire rows receiving an estimated UV-C dose of 0 J/m2, 80 J/m2 or 120 J/m2 once a week in 2020, and 0 J/m2, 120 J/m2 or 200 J/m2 twice a week in 2021. UV-C dose was adjusted using tractor ground speed.  The five, 5-vine subplots within each row are managed with different fungicide programs. In 2020, subplots received an untreated control, 7 or 14-day interval sulfur (Microthiol; 5 lbs/A), 14-day interval sulfur (5 lbs/A), 14-day interval QoI (Abound Azoxystrobin; 10 floz/A), and in 2021 subplots received an untreated control, 14-day interval sulfur (Microthiol; 2 lbs/A), 14-day interval sulfur (4 lbs/A), 21-day sulfur (4 lbs/A), and an alternation on a 14-day interval of sulfur (4 lb/A) and a QoI (Abound Azoxystrobin; 10 floz/A). In 2021, treatment rows were separated by buffer rows treated with 14-day interval sulfur (4 lb/A) and either 120 or 200 J/m2 UV-C applied once a week. UV-C and Fungicide treatments started at 6-inch shoot growth and ended at veraison (grape color change).

Powdery mildew leaf incidence was assessed every other week starting at 12-inch shoot growth and ending at veraison by randomly selecting ten leaves from each of the three center vines within each subplot (30 leaves rated per subplot). Swab samples were also taken of powdery mildew colonies found by raters by touching a cotton swab to discover colonies. In addition, gloves of the disease raters were swabbed, rinsed with water, and dried after rating each subplot. Cotton swabs were processed at the lab to extract DNA from the conidia collected and quantify the number of E. necator conidia (Thiessen et al. 2016) and the presence of alleles associated with fungicide resistance (Miles et al. 2020). At the start of veraison, severity ratings of mildew on 10 leaves and 10 clusters per vine were visually assessed from the middle three vines of each subplot. For leaves, the area of both the adaxial and abaxial sides was estimated using a reference scale. For clusters, the percent area of the cluster with mildew infection was estimated using a reference scale.

To rate for bunch rot, 5 clusters from each of the middle three vines (15 total) of each subplot were randomly sampled at harvest, brought back to the lab, rated for bunch rot severity (Sanzani et al. 2012), then each cluster was placed in plastic planter tray cells, wrapped in a plastic trash bag, and put into a 20°C growth chamber and incubated in high humidity (85%) for 48 hours and rated again. After the incubation, cotton swabs were touched to sporulating Botrytis colonies, washed into 0.05% Tween 20, centrifuged, and conidia resuspended in 25% glycerol and stored at -20°C to be later cultured for fungicide resistance assays.

All data were analyzed using RStudio to compare treatments for differences in disease development and frequency of fungicide resistance.

Objective 3. Examine the suitability of an autonomous robotic platform to deliver UV-C applications at a commercial scale.

Willamette Valley Vineyards purchased a SAGA robotics UV-C unit which was delivered and assembled in April of 2021. Tests of the unit’s software, hardware, mobility, and autonomous platform were conducted on vineyard rows at WVV and the Oregon State University Botany and Plant Pathology vineyard. Deployment for field trials was attempted but issues were discovered limiting the capabilities to execute trials reliably and safely.

UV-C array robot moving down a vineyard row
Figure 3. The autonomous SAGA robotics UV-C platform being tested at Willamette Valley Vineyards.

Objective 4. Determine the impact of UV-C application on fruit chemistry

UV-C effects on fruit chemistry. At harvest, an additional 15 harvest-ready clusters were randomly sampled from the three center vines of each five-vine subplot at the BPP site. Samples were stored at -80°C until they could be processed. Berries from clusters were randomly subsampled into three, 100-gram samples and thawed before being homogenized in a Waring blender on the highest setting for 2 minutes. Berry homogenate was centrifuged at 1730 x g for 10 minutes and the supernatant collected calculated soluble solids as degrees Brix with a refractometer, pH measured by a pH meter. Fruit anthocyanins and phenols were measured by acidic ethanol extraction of berry homogenate supernatant and were estimated using spectrophotometry to compare the no UV-C control treatments to the UV-C treated fruit (Bindon et al. 2014; Mercurio et al. 2007; Rantsiou et al. 2020). Samples for fruit chemistry were not taken of the untreated control vines in 2020 since the disease desiccated the berries and was not representative of realistic vineyard management. 

Research results and discussion:

Lab Studies

For isolates of Erysiphe necator, the preliminary effective 50% inhibitory dose (ED50) of UV-C values varied across isolates from 118 to 259 J/m2 with an average of 172 J/m2 (Table 1). This assay measures the germination growth of the conidia, and not the pathogenicity or true viability of the conidia, so the ED50 to inhibit infection may be lower than reported here. Regardless, this assay can still show the relative differences across isolates. This variability in sensitivity may suggest that there may be cellular mechanisms that influence the tolerance or sensitivity of powdery mildew to UV-C.  More powdery mildew isolates will be tested in the future and repeated experiments on isolates tested to date. There has not seemed to be a pattern of tolerance to UV-C and a tolerance to fungicides such as Quinone inhibitors (QoIs). For Botrytis, conidia were considerably more tolerant to UV-C than the 24-hour germlings (Figure 4). This may be due to the higher degree of melanin in conidia than in the growing germination tube, allowing the conidia to be more tolerant to UV-C. This suggests that any control strategies using UV-C to manage Botrytis should be focused on periods suitable for infection to target germinating spores or mycelia growing saprophytically on host tissue. 

Table 1. Preliminary UV-C ED50 estimates to inhibit E. necator conidia

Isolate name




































Graph showing the dose response curve of Botrytis conidia and germlings to UV-C
Figure 4. Dose-response curve of Botrytis conidia (blue) and 24-hour old germlings (red) to UV-C. Each point represents a plate well relative growth to the control. Shaded areas represent the standard error of the mean.

Field Trials

In field trials, once weekly UV-C application in 2020 led to a significant reduction in the incidence of powdery mildew on leaves expressed as the area under the disease progress curve (AUDPC was reduced by 28% and 17% for 80 and 120 J/m2 applications, respectively (Figure 5A)). Neither the incidence of Botrytis bunch rot on clusters nor the frequency of QoI resistant E. necator detected was not reduced by once-a-week UV-C treatments.

In 2021 field trials, twice-weekly UV-C application on average significantly reduced the leaf incidence AUDPC, reducing the AUDPC by 53% and 56% for 120 and 200 J/m2 applications, respectively (Figure 5B) and mildew cluster incidence by 22% and 35%, respectively (Figure 6). However, once weekly applications of UV-C in conjunction with a 14-day interval, micronized sulfur fungicide spray program was not significantly different from the twice-weekly applications of UV-C with the same fungicide program. UV-C application did not significantly reduce Botrytis incidence on clusters and did not reduce the frequency of QoI resistant powdery mildew. Using glove swabs, (Thiessen et al. 2016) the E. necator inoculum detected (log-transformed) from glove swabs over the field season (expressed as area under the curve) was reduced by 24% and 37% with twice weekly applications of 120 and 200 J/m2, respectively, compared to non-UV-C treated plots (Figure 7). This indicates that in addition to there being a reduced epidemic seen in visual scouting, powdery mildew inoculum levels in the plots may have also been reduced by UV-C application.

Bar graph showing the reduction in powdery mildew leaf disease by UV-C in conjunction with fungicides
Figure 5. The area under the disease progress values (units arbitrary) for powdery mildew leaf incidence epidemic in 2020 (A) and 2021 (B) with once or twice weekly UV-C applications, respectively, in conjunction with fungicide spray programs. Error bars represent the standard error of the mean.
Bar graph showing the reduction of cluster powdery mildew incidence in 2021 with UV-C application
Figure 6. Powdery mildew cluster incidence epidemic in 2021 with twice-weekly UV-C applications, respectively, in conjunction with fungicide spray programs. Error bars represent the standard error of the mean.
Bar graph showing the reduction in detected inoculum with UV-C application
Figure 7. Powdery mildew inoculum detection from rater glove swabs in 2021. Bars represent the area under the curve (units arbitrary) for the log-transformed detected conidia over the season with twice weekly UV-C in conjunction with fungicide spray programs. Error bars represent the standard error of the mean.

Robotic UV-C array

The robotic array was found to not be robust enough for vineyard use in its current development. Issues discovered included: not being water and dust tight, difficulty with slopes (up or cross slopes), and common loss of wheel alignment due to rough terrain. Future directions will be focused on optimizing and improving the robotic array as well as using the Monarch electric tractors as the delivery system. Willamette Valley Vineyards has an order for a Monarch tractor and a new lighter-weight UV-C tractor-mounted unit is planned to be built for the electric autonomous Monarch tractor system.

Fruit Chemistry

Brix content, pH, anthocyanin, and phenolic levels of whole grape berry homogenates were not significantly different across the UV-C treatments for fruit from 2020 and 2021 field trials. This indicates that the UV-C treatments will not be detrimental to the fruit quality, and in turn, crop value. 


These results suggest that UV-C treatments, in conjunction with fungicide programs and other integrated pest management strategies, may have the potential to allow growers to use fewer fungicides or increase spray intervals between applications. Reducing fungicide applications would save growers on input costs associated with fungicide applications, and the environmental inputs from fungicide applications. Even though UV-C offers promising improvement in powdery mildew disease control, the application of UV-C must be made more economically feasible for commercial use, thus future work will continue to examine the use of autonomous robotic platforms for the application of UV-C.

Participation Summary
1 Farmer participating in research

Educational & Outreach Activities

10 Consultations
3 On-farm demonstrations
5 Published press articles, newsletters
2 Tours
10 Webinars / talks / presentations
2 Workshop field days

Participation Summary:

250 Farmers
250 Ag professionals participated
Education/outreach description:

Alex presented UV-C research at the following events:

  • American Phytopathological Society Pacific Division meeting 2021: Virtual presentations by the graduate students showcasing their research. Most participants were academic researchers.
  • OWRI Spring webinar 2021: A virtual seminar series focusing on the research of the members of the Oregon Wine Research Institute with 25 people attending the virtual seminar session.
  • Sustainable Agriculture Expo 2021: The Sustainable Ag Expo is a 6-week of virtual learning seminar and tradeshow that provides an opportunity for farmers, ag professionals, and pest control advisors to learn about the latest in farming research, resource issues, and business trends related to sustainable agriculture. Of the 121 participants in the course, 101 completed the course and took a quiz on the content, and passed.
  • Grape Day 2022: Grape Day is an annual event run by the Oregon Wine Research Institute at Oregon State University highlighting research relevant to the Oregon wine industry. The goal is to provide industry members with an opportunity to network and discuss research with our researchers and invited speakers. There were 118 participants, with a mix of growers, ag professionals, and researchers. 
  • Vineyard Technology Field Day 2022: An annual event run by the OSU Viticulture Extension, USDA-ARS, and Willamette Valley Vineyards to demo emerging technologies and innovations in disease management, vineyard nutrient monitoring, crop estimation, and alternative-powered tractors. There were 100 participants, most of which were growers. 
  • International Workshop on Grapevine Downy and Powdery Mildews: An international workshop on recent research on powdery and downy mildew of grapevines. There were approximately 75 participants, comprised of mostly researchers.

Alex and Walt also published research on UV-C for disease management in the Vine to Wine OWRI extension publication in late 2020 and the research was featured in The Furrow magazine in November 2021.

In addition, Walt Mahaffee presented the UV-C research at the following events:

  • LIVE annual meeting. Comprised of ~57 growers and ~10 ag professionals
  • Sonoma Grape Growers Association meeting with 186 in attendance
  • Costal Viticulture Consultants meeting with 43 in attendance
  • Easter Oregon Growers meeting with 57 in attendance

Walt has also provided information, consultations, and/or tours on the UV-C array design and research results to ten individuals. 

Willamette Valley Vineyards also held two of their own field days with growers, and local TV stations covering the events. WVV has also published press releases to show the UV-C array technology being used on site. 

Project Outcomes

21 Farmers reporting change in knowledge, attitudes, skills and/or awareness
8 Farmers intend/plan to change their practice(s)
1 Grant received that built upon this project
Did this project contribute to a larger project?:
2 New working collaborations
Project outcomes:

Currently, UV-C technology in the Willamette Valley, Oregon has shown potential as an integrative management tool to help reduce grape powdery mildew. Our small plot work has shown that scaling this technology up to commercial scale testing is justified. The results suggest that UV-C applications might allow growers to use less fungicides by reducing fungicide rates and extending intervals between sprays, reducing the economic, and environmental costs associated with managing powdery mildew. However, it is not clear that current commercial acceptable disease levels can be achieved at commercial production scales. With consumer demands trending toward products that are certified to be sustainable, this technology could potentially help growers more easily meet sustainable certificate standards. However, this technology needs to be made cost-effective reducing the labor requirements (e.g., used on electric autonomous platforms) to become a feasible management option.

Knowledge Gained:

Alex: During the time of this grant, I found that sustainable agriculture is going to be reliant on making the tools that we develop for growers should be accessible to all growers who want them. Each grower's site has its own limitations and constraints which do not allow these tools, such as UV-C, to be a one size fits all solution. Making these UV-C array designs open source, and accessible in the construction and implementation, will promote adoption by allowing growers to customize the tools for their unique situations.

Walt: Reaffirmed that most grape growers are always looking to improve the sustainability of their production system but are also limited by the time and the successful implementation efforts need to address this time component.  The project also reaffirmed that it is better to teach how to achieve goals that provide specific solutions.  Growers will adapt and mix and match.  In this case, demonstrating multiple ways of delivering UVC empowers growers to adapt the technology to their needs and equipment.

Information Products

    Any opinions, findings, conclusions, or recommendations expressed in this publication are those of the author(s) and do not necessarily reflect the view of the U.S. Department of Agriculture or SARE.