Battling Fire Blight with Biologicals

This post was written by Anna Wallis, Kerik Cox, and Mei-Wah Choi (all from Cornell’s School of Integrative Plant Science, section of Plant Pathology and Plant-Microbe Biology). Thanks for sharing your research with us!

Since this is a slightly longer post, here’s a little table of contents:

Biological Modes of Action

What products are currently available and where do they fit in?

Results from the Cox lab

The verdict on biologicals for fire blight management

Streptomycin is a clear asset in the fire blight arsenal—it is inexpensive, effective, and reliable. However, antibiotics may not always be a viable option. More and more, biological materials are holding their own in the fight, with an increasing number of products on the market claiming protection for both blossom and shoot blight. Biological materials are still relatively new to the apple scene, an industry with a long track record of effective disease management. So why change to biologicals, and how do they work?

There are a multitude of reasons driving the growth of antibiotic alternatives. Organic production eliminated antibiotic use in 2014 in the United States. In European markets, they are prohibited or severely limited. Pressure from regulatory organizations and markets to use more sustainable management techniques will not be slowing any time soon. The prevailing evidence supports that responsible streptomycin applications do not seem to select for resistance in the pathogen. Yet, resistance continues to appear in commercial settings.

So, what are these biological materials and how do they work? In the ‘What is Biocontrol?’ tab above, Amara provides an excellent overview of biocontrol, as defined by the EPA and industry. Here I’ll review the biological modes of action and specific materials available in the context of fire blight management. I’ll also provide a snapshot of how biological programs have performed in our research orchards. There is no intention to endorse any specific trade products, rather this is an attempt to provide a neutral perspective and overview of the current market.

Biological Modes of Action

Biological materials available for fire blight management are typically biopesticides falling into the biochemical or microbial category. This means they are derived from natural sources (i.e. plant extracts or minerals) or they are composed of microcorganisms and/or their products.

To understand how biologicals can be used in fire blight management, it’s first important to review the important features of the disease. A thorough description of the disease cycle, symptoms, and causal organism can be found on this Cornell Fact Sheet. Fire blight is caused by Erwinia amylovora, a bacterial pathogen which preferentially colonizes the floral surface, specifically the stigma or the sticky part of the tip of the female organ. First, enough heat must be accumulated for colonization to occur, which can be predicted by disease forecasting models such as MaryBlyt (if you’re familiar with the disease and pest prediction tool NEWA, this is the model used in the fire blight prediction model there). Then there must be a wetting event to wash the bacteria into the natural openings in the flower, the nectary at the base of the floral cup. Unlike fungi, bacteria cannot penetrate plant cells directly, so they rely on natural openings and tissue damage to invade their host.

Pictures (counterclockwise, from top left) of a yellow glob of bacteria oozing out of a dark, necrotic canker on an apple stem; a dead cluster of apple blossoms; a dead apple shoot curved like a shepherd’s crook; the base of an apple tree trunk with a dark canker.
Figure 1. Simplified disease cycle for Erwinia amylovora, causal agent of fire blight. Clockwise from top left: primary inoculum is produced in the spring as bacterial ooze from old cankers; inoculum is transferred to open flowers and causes blossom blight; blighted blossoms provide additional inoculum which is transferred to young leaf tissue damaged by wind or hail causing shoot blight; bacteria may also travel systemically via the vascular system of the plant leading to canker blight; cankers produced from blossom, shoot, or canker blight provide an overwintering site for bacteria to colonize the tree in the following season.

Biologicals can disrupt these events by:

  1. Outcompeting the bacteria during colonization of the plant
  2. Producing antibiotic metabolites, killing the pathogen prior to infection, or
  3. Priming natural host defenses, making the plant more resistant to the bacteria. This is called ‘Induced Resistance’

A simplified view of these events is depicted in Figure 2.

Two identical pictures of a cluster of open apple blossoms. Left photo with a red circle around the yellow floral parts (stigmas) and a blue curved arrow from this circle to the base of the flower (nectary), representing the colonization of the stigma by E. amylovora and a wetting event washing the bacteria into the plant. Right photo with the red circle and blue arrow, plus a yellow ‘X’ over the red circle, indicating protectant activity at the stigmatic surface, and a brown ‘T’ with the top facing the stigma, indicating the induction of plant defenses.
Figure 2. Depiction of fire blight blossom infection and how biological materials interfere. (A) In order for a blossom infection to occur, flowers must be open and receptive, heat accumulation must be sufficient for E. amylovora to colonize the stigma (red circle), and there must be a wetting event (blue arrow) to wash the bacteria into the floral nectary. (B) Biological materials protect against infections by outcompeting the pathogen or producing antibiotic metabolites (yellow ‘x’) or priming host defenses (red letter “T”).

Like any product, these materials require precise applications, to ensure they are in the right place at the right time to provide effective control (Figure 3). Materials with competitive action or antimicrobial metabolites that ‘protect’ the flower (protectants) must be applied when the bacteria is present or just before. This enables sufficient, timely colonization or interaction with the pathogen. Induced resistance materials (defense inducers), also called Systemic Acquired Resistance or Induced Systemic Resistance materials (SARs or ISRs), must be applied prior to infection events, with enough time to activate the host response. (Click the image below to enlarge it.)

Seven pictures of apple buds in a row, depicting growth stages in chronological order: 1. dormant (closed, brown buds), 2. green tip (green leaves just starting to emerge), 3. half inch green (about ½” of green showing), 4. tight cluster (cluster of green floral buds in center of emerged leaves), 5. pink (floral buds showing pink color), 6. bloom (open blossoms), and 7. petal fall (cluster of very small fruitlets). At several stages, words are printed above indicating actions to be taken for fire blight management. At dormant “Prune out cankers”, at green tip “Copper”, at pink “Pre-bloom defense inducers”, at bloom “Protectants”, at petal fall “Post-bloom defense inducers.”
Figure 3. Approximate timing of biological materials corresponding to phenological stages of apple for blossom and shoot blight protection. In any fire blight management program, it is essential to remove inoculum (old cankers) during the dormant period and apply a general antimicrobial at green tip to reduce inoculum. Blossom blight control is provided by defense inducers applied prior to bloom and protectants applied at bloom. Additional applications of defense inducers post-bloom provide shoot blight control; some of the earlier applications targeting blossom blight seem to also have some carry-over effect for shoot blight.

What products are currently available and where do they fit in?


Blossom protectant type products include both bacteria and fungi. The most well-known examples include: Pantoea agglomerans, a bacterium closely related to the fire blight bacterium and an excellent colonizer of apple flowers, marketed as Bloomtime Biological (Northwest Agricultural Products), and the yeast Aureobasidium pullulans, a fungus, marketed as Blossom Protect (Westbridge Agricultural Products). Another bacterium, Pseudomonas fluorescens, is also an effective competitor and is marketed as BlightBan (NuFarm).

Materials with antimicrobial activity are most often Bacillus species, most commonly strains of B. amyloliquefaciens and B. subtillus. Currently on the market are Serenade Optimum (Bayer), Double Nickel (Certis), and Serifel (BASF).

Products that stimulate Induced Resistance response in the host plant work by stimulating two possible pathways the ISR and SAR, as mentioned earlier. These pathways are related and overlapping in the plant, and scientists are still detangling the complex molecular mechanisms involved in plant protection. Example products include Regalia, an extract of the plant Reynoutria sachaliensis or giant knotweed (Marrone Bio Innovations) and a Bacillus mycoides strain marketed as LifeGard (Certis). Another common product used in induced defense is acibenzolar-S-methyl. This is not a biological, but a synthetically derived product marketed as Actigard (Syngenta).

Many of these products have been recommended as part of an integrative management strategy outlined in an extensive report from The Organic Center, based on results from both research trials and anecdotal experience (Ostenson and Granatstein 2013). Always follow the label on any pesticide (including biopesticides) you use.

Table 1. Biological products for Fire Blight

Product Active Ingredient Mode of Action
Firewall Streptomycin antibiotic – kills pathogen
Blossom Protect Aureobasidium pullulans strains DSM14940 & 14941 competitive with pathogen
Bloomtime Biological Pantoea agglomerans strain E325 competitive with pathogen
BlightBan Pseudomonas fluorescens strain A506 competitive with pathogen
Serenade Optimum Bacillus amyloliquefaciens strain QST713 antibiotic metabolites
Double Nickel Bacillus amyloliquefaciens strain D747 antibiotic metabolites
Serifel Bacillus amyloliquefaciens strain MBI600 antibiotic metabolites
Regalia extract of Reynoutria (giant knotweed) resistance inducer
LifeGard Bacillus mycoides isolate J resistance inducer

Results from the Cox lab


Our lab conducts extensive trials evaluating efficacy and sustainability of disease management programs in our research orchards at Cornell AgriTech in Geneva. More recently testing has included various biological materials. In these trials, management programs are tested in two orchard blocks: a Gala block and an Ida Red block, established in 2002 and 2004 respectively, both on B.9 rootstock. The trees in these blocks are spaced considerably farther apart than commercial orchards in order to prevent drift between treatments.

Programs targeted either blossom or shoot blight. To provide sufficient disease pressure, trees are inoculated with a high concentration of E. amylovora at bloom. In blossom blight programs, resistance inducers are applied at pink, and protectants are applied at bloom. For shoot blight programs, resistance inducers are applied at petal fall.

Disease pressure varied from season to season, as indicated by the untreated control trees, ranging from 60 to 99 % disease incidence. Across all trials, antibiotics provided the most consistent and reliable control of both blossom and shoot blight, with less than 15% blossom and 5% shoot blight. The biological materials, both protectants applied at bloom and defense inducers applied pre-infection, also provided good disease protection with typically less than 30% incidence depending on the season conditions and the product. Compared to antibiotic programs, these materials showed greater variation both within and between seasons (i.e. greater standard deviation within a treatment and different top performers in different seasons). In seasons with lower disease pressure, biological programs tended to perform as well as antibiotics. Some of the specific results from 2015-17 are shown in Figure 4 (click the image to enlarge the graphs).

Disease was most severe in the untreated control, ranging from 60% to more than 90% of blossoms blighted and 30% to more than 50% of shoots blighted. Pressure was high in 2015 and 2017, lower in 2016. The antibiotic streptomycin always had <20% and often 0% incidence. Defense inducers outperformed protectants in 2015. In 2016 and 2017 defense inducers and protectants performed similarly, and overall disease incidence was lower.
Figure 4. Average disease incidence of four replicate trees treated with fire blight management programs in 2015 (A & D), 2016 (B & E), and 2017 (C & F). Programs included untreated control (grey bar; highest disease pressure), antibiotics (maroon), resistance inducers (blue), and blossom protectants (yellow).

The verdict on biologicals for fire blight management


Do we recommend biological materials for fire blight management? Overall, the answer is generally yes. There are several important considerations to consider. In our research orchards, the system is challenged with a very high level of inoculum to examine fine differences in product performance. These inoculum levels are much higher than would be present in most commercial orchards. Hence, we expect all programs would perform even better in a commercial setting. In addition, combinations of products seem to be the best: for example, pairing a defense inducer applied at bloom with a protectant material at bloom to control blossom blight, with follow up defense inducer applications for shoot blight. We also expect efficacy of biological materials to improve in the future. Changes in formulations improving activity (note the old and new Regalia formulations in Figure 3), as well as shelf life, tank mixing, and storage happen fairly regularly and will make products more accessible and affordable for growers.

Biologicals are still relatively new materials. As with any product, there is still much to learn about how products work in the field, the most effective management programs, and translating best practices from research to commercial settings. We believe they are a valuable part of an integrated fire blight management approach, including good cultural and mechanical practices such as planting resistant cultivars and rootstocks and removing inoculum from the orchard.

 

You can learn more from these sources:

Ostenson, H., and Granatstein, D. Grower Lessons and Emerging Research for Developing an Integrated Non-Antibiotic Fire Blight Control Program in Organic Fruit. The Organic Center. November 2013. Available at: https://www.organic-center.org/wp-content/uploads/2013/07/TOC_Report_Blight_2b.pdf

Pal, K., and Gardener, B. 2011. Biological Control of Plant Pathogens. The Plant Health Instructor, APS. Available at: https://www.apsnet.org/edcenter/advanced/topics/Pages/BiologicalControl.aspx.

Turechek, W. W., and Biggs, A. R. 2015. Maryblyt v. 7.1 for Windows: An Improved Fire Blight Forecasting Program for Apples and Pears. Plant Health Progress. 16:16–22. Available at: https://www.plantmanagementnetwork.org/pub/php/volume16/number1/PHP-RS-14-0046.pdf

And the results are in…from Year 1. What do biofungicides add to vegetable disease management Part 3

Cucurbit powdery mildew on a winter squash leaf.
One of our goals for this project was to understand what biofungicides might add to a cucurbit powdery mildew management program.

Introduction

In 2018 we conducted field trials using biofungicides in cucurbit powdery mildew and snap bean white mold management programs. Hopefully you’ve read part 1 and part 2 of this biofungicide story. If not, now might be a good time.

Part 1 will give you more details about the trial design. We wanted to know whether adding biofungicides would improve disease control, plant health, or yield. For cucurbit powdery mildew, we were adding one of three different biofungicides to a conventional chemical spray program. We also included a treatment that was all OMRI-listed (organic) products. For white mold on snap beans, we were curious about using an in-season biofungicide (Double Nickel, Bacillus amyloliquefaciens strain D747) in combination with a pre-season biofungicide (Contans, Paraconiothyrium minitans strain CON/M/91-08). In 2018, our white mold treatments were just Double Nickel and Cueva (an OMRI-listed copper). In 2019, we’ll add the pre-season Contans treatment.

Part 2 explains more about the modes of action of the five biofungicides we are looking at. The post also includes practical information about how to use these biofungicides to maximize their efficacy – compatibility with other products, best way to store them, when to apply them, etc.

Now it’s time to talk about what we learned from this first year (of a two-year project).

The bottom line

We don’t want to keep you in suspense, so here’s a quick summary of what we learned. Fortunately for the eastern NY grower who graciously allowed us to run the trial on his farm (but unfortunately for us), the snap bean field had very little white mold in 2018. Even the plots that were not sprayed with Double Nickel or Cueva had almost no disease. So we weren’t surprised when there were no differences in disease, plant health, or yield among the white mold treatments. Results from Sarah Pethybridge’s efficacy trials with OMRI-approved products for white mold are available online.

A healthy field of snap bean plants.
Our snap bean trial in eastern NY in 2018 had very little white mold. (Photo credit: Crystal Stewart)

Cucurbit powdery mildew was a bit more severe than white mold (low pressure in eastern NY, moderate pressure in western NY and on Long Island), but we were not able to detect statistically significant benefits from adding biofungicides to a conventional spray program. Disease severity, plant health (as measured by NDVI), yield, and fruit quality (Brix) were the same whether you used a conventional spray program, or a conventional spray program plus a biofungicide. We didn’t measure significant differences in yield among any of the treatments at any of the three sites.

The conventional powdery mildew spray program alone, or when combined with LifeGard, Regalia, or Serifel significantly reduced disease compared to no treatment for cucurbit powdery mildew. Adding any of the biofungicides to the conventional spray program did not improve control compared to using only the conventional sprays. The organic (OMRI-listed products) treatment was not significantly different from either no sprays at all, or the conventional spray program.
Severity of powdery mildew on the upper sides of the leaves in the Western NY trial. Here, disease severity is quantified using the area under the disease progress curve (AUDPC). This number summarizes disease severity from multiple dates, and the larger the number, the worse the disease. If two treatments share the same letter, the average disease in those treatments is not significantly different. The error bars give you an idea of how much variability there was in each treatment.

NDVI results

NDVI (normalized difference vegetation index) values did not detect cucurbit powdery mildew early. (Since there was so little white mold, we couldn’t test NDVI for early detection.) There was some inconsistent correlation between NDVI readings and disease, yield, and Brix in winter squash. In WNY we used both a handheld GreenSeeker and a gator-mounted Crop Circle to measure NDVI. Both devices had similar results. Based on this first year of testing with these two devices, NDVI measurements were not useful as an early indicator of cucurbit powdery mildew.

In addition,  NDVI measurements did not  detect subtle differences in plant health among treatments. At only one of our three sites (Long Island) were there any significant differences in NDVI among treatments. This was only on the last two rating dates in the season, when powdery mildew was visibly more severe in the non-treated control than the conventional fungicide treatments.

On the last two rating dates of the season (August 31 and September 17), NDVI values were significantly higher in the conventional powdery mildew spray program treatment and all three of the conventional + biofungicide treatments, compared to the plots that were not treated for powdery mildew. Adding the biofungicides did not significantly improve NDVI, compared to using only conventional products.
Normalized difference vegetation index (NDVI) measured on winter squash in the Long Island trial on three dates at the end of the season. NDVI values closer to 1 indicate more healthy, green foliage. If two treatments have the same a letter on the same date, the average NDVI readings on that date were not different between the two treatments. Data from August 31 are labeled with uppercase, while data from September 17 are labeled with lowercase letters. There were no differences among any treatments on August 24. The error bars give you an idea of how much variability there was in each treatment. We couldn’t do statistics on the organic treatment because too many plants were killed by Phytophthora blight in the plots that received this treatment.

Some caveats

The non-treated control (received no powdery mildew fungicide) was often not significantly different from the conventional fungicide control (our best management program). We know that controlling powdery mildew on cucurbits is important, so if we don’t detect a significant difference between the non-treated control and the treatment that should have provided the best control, it is then hard to draw further conclusions from the data.

We didn’t measure statistically significant differences in marketable yield among any of the treatments at any of the sites. Data for eastern NY are shown in this graph.
We didn’t detect statistically significant differences in marketable yield among any of the treatments in any of the trials. Here are the data from eastern NY. Notice that all six bars are labeled with the letter “a”. As with previous graphs, the error bars give you an idea of how much variability there was across the different plots in each treatment.

When disease pressure is low (as it was in Eastern NY), we would expect not to see many differences between treatments. Similarly, if the conventional fungicide program provided excellent disease control (as it did on Long Island), it would be hard to detect an improvement in control from adding a biofungicide. Another challenge we dealt with in the Long Island trial was Phytophthora blight. By the end of the season, we had lost two of the four plots receiving the organic treatment to this disease. This limited our ability to statistically analyze the biofungicide data. On Long Island, the organic spray program initially performed well – as seen on August 31  – comparable to the conventional treatments. But by the final assessment on September 17, the organic program was no longer as effective. This was not surprising since it was 10 days after the last application. Suffoil-X was the final organic product applied, and it has little residual activity.

In the Long Island trial, there was very little disease on August 16 or 24. On August 31, disease had increased in the non-treated control, but the organic treatment was still suppressing disease well. Control in the organic treatment had declined by September 17, but this was 10 days after the last spray was applied.
Average severity of powdery mildew on the upper surface of leaves on the last four assessment dates in the 2018 Long Island trial. All of the treatments (except the non-treated control) suppressed powdery mildew well through August 31. Control in the organic treatment had declined by September 17, but this was 10 days after the last spray was applied.

In WNY, we had an epic aphid outbreak. An entomologist colleague identified them as probably melon aphids, and also that 2018 was generally a bad year for aphids. It’s also possible that while trying to control cucumber beetles earlier in the season, we killed some aphid natural enemies, contributing to an aphid outbreak later in the season. I know cucumber beetles are tough, but if you can manage them without decimating your local natural enemies, you’ll be doing yourself a favor!

The underside of a squash leaf covered with aphids; an acorn squash fruit covered with shiny honeydew from aphids; a close-up picture of an adult aphid and some young aphids.
The severe aphid outbreak in the western NY trial may have made it more difficult to detect differences among treatments. In late August, some of the leaves were covered with aphids (A), and many fruit were covered with honeydew (B). Getting a close look at the aphids is essential for correct identification (C).

We deliberately used a very intensive spray program, starting our biofungicide applications early, and continuing to apply them as we added conventional fungicides later in the season. This was an expensive powdery mildew management program. But, in this first year of the project, we didn’t want to be left wondering if a lack of differences was due to underapplication of the biofungicides.

If you want to see more of the data we collected from the cucurbit powdery mildew trial, you can find it in the Proceedings from the 2019 Empire State Producers Expo.

What does this all mean?

First, this is only the first year of our project and one year of data. It’s a start, but we’ll hopefully learn more in a second year. Since we didn’t measure a significant improvement in yield, we didn’t see evidence that adding biofungicides to a full chemical spray program for powdery mildew justified the cost. The relative costs of the treatments we used are listed in the table below, and the approximate per acre costs of each product are in the Proceedings from the 2019 Empire State Producers Expo. Replacing a chemical spray or two with a biofungicide could be a more economical option. That’s something we’re planning to look at in 2019.

Treatment
Date Non-treated Conventional Conventional + LifeGard Conventional + Regalia Conventional + Serifel Organic
7/19/18 LifeGard Regalia Serifel LifeGard
7/27/18 LifeGard Regalia Serifel LifeGard
8/3/18 Vivando LifeGard + Vivando Regalia + Vivando Serifel + Vivando MilStop
8/10/18 Quintec LifeGard + Quintec Regalia + Quintec Serifel + Quintec Serifel
8/17/18 Luna Experience LifeGard + Luna Regalia + Luna Serifel + Luna SuffoilX
8/24/18 Vivando LifeGard + Vivando Regalia + Vivando Serifel + Vivando MilStop
8/31/18 Quintec LifeGard + Quintec Regalia + Quintec Serifel + Quintec Serifel
9/7/18 Luna Experience LifeGard + Luna Regalia + Luna Serifel + Luna SuffoilX
Total cost (per A) $228.28 $343.32 $536.28 $696.28 $257.76
Cost increase vs. conventional (per A) $  – $115.04 $308.00 $468.00 $29.48

Based on results from this year, we can’t yet recommend that you run out and buy a handheld NDVI sensor for early detection of cucurbit powdery mildew. We’ll collect NDVI data again in 2019, and let you know what we learn. Although our results from the field trials were somewhat inconclusive in this first year, we’re hopeful that the information we’ve compiled about how these biofungicides work and how to use them will be useful. If you’re thinking of using Contans, Double Nickel, LifeGard, Regalia, or Serifel in 2019, first take a look at these fact sheets related to our white mold and powdery mildew trials. And if you have used biofungicides, we’d be interested in hearing about it; click here to send an e-mail.

This post was written by Amara Dunn (NYS IPM), Elizabeth Buck (Cornell Vegetable Program), Meg McGrath and Sarah Pethybridge (both Plant Pathology & Plant-Microbe Biology, School of Integrative Plant Science, Cornell University), Crystal Stewart (Eastern NY Commercial Horticulture Program), and Darcy Telenko (Department of Botany & Plant Pathology, Purdue University). Thank you to the New York Farm Viability Institute for funding.