Can frost events trigger fire blight? Frost injury is considered a fire blight “trauma event”, but very little is known or published about exactly what conditions are required to trigger fire blight during a frost event. Because frost was widespread throughout New York and New England on the morning of May 18, there is at least a slight risk that trauma blight may be appearing in some orchards next week. However, frost-induced fire blight is so rare that apparently no one thought about it on the days just before or after the frost when a strep spray might have been useful. However, as outlined below, even if I had thought of the potential for frost-induced trauma blight, I probably would not have recommended applying strep because the the late timing and dry weather associated with this year’s frost event may have negated most of the risk.
Frost can create an infection event for Erwinia amylovora because frost injury provides entry points and because melting ice crystals in fruitlets and shoot tips create a slight “suction” as they melt, thereby increasing the likelihood that bacteria will be drawn into the injured tissue. Limiting factors for frost-induce fire blight presumably involve presence/absence of abundant inoculum on the surface of host tissues at the time of the frost as well as other poorly understood effects of the environmental conditions before and after the frost event.
Over the course of my career, I have seen at least one, and perhaps two, devastating fire blight outbreaks in apples related to frost trauma. The first occurred in the early 1980’s south of Gardiner, NY, at the time when Dr. Paul Steiner was developing his MaryBlyt model. I was sending Paul data to use in building his model, and the outbreak near Gardiner did not fit any scenarios for blossom blight infection. However, Paul noted that a frost event had occurred, and he used the timing of symptom appearance to trace the infection back to that frost event (which occurred during bloom, as I recall). I saw the second probable frost-related blight outbreak in Quebec about a decade later. In both cases, fire blight was so extensive that it could not be pruned out and in some cases, orchards were removed due to severity of the fire blight later that season.
After spending several hours looking at the weather data and fire blight predictions for numerous locations on the NEWA website, I can summarize as follows:
- Most locations in the Hudson Valley and southern New England were registering EIP’s of 100 to 250 on May 16, but the EIP levels dropped by about a third on May 17 due to cooler weather. An EIP >100 is required to trigger blossom blight when moisture is present, but no one knows whether enough inoculum may be present even at lower EIP levels to allow infection via frost injury. The NEWA EIP model showed only high risk, not infection, for fire blight on May 15-16 because weather was dry. Risk levels for Cougar Blight were more elevated.
- Note that the program used to calculate EIP zeros out the EIP number after a frost. However, that does not mean that the risk of frost-induced trauma blight was also zero at the time of the frost. Bacteria not drawn into frost-damage tissue apparently do not survive frosts well enough to be of concern after the event, but the potential for frost-induced trauma blight should be judged based on the EIP the day before the frost. As noted above, we don’t know if bacteria can trigger frost-induced infections even when EIP values are below 100 at the time of the frost. However, I suspect that risks of frost-induced trauma blight are rather low when the EIP is below 100 at the time of the frost (in this case, on May 17), although presence of oozing cankers may override the importance of EIP during frost events.
- In the lower Hudson Valley and southern New England, most trees were well past petal fall by May 18. The fire blight bacterium multiplies primarily on the stigma of flowers, and after bloom that site is no longer available. Thus, inoculum levels at the time of frost on May 18 should have been quite low EXCEPT THAT ….
- Orchards with active cankers from previous years might have had high inoculum levels on shoot tips and fruitlets due to dissemination of bacteria from oozing cankers by various insects. Also, newly planted orchards and cider apple cultivars may still have had open flowers on May 18, and those could also have contributed inoculum for frost-induce trauma blight.
- Orchards with actively oozing cankers AND open flowers were almost certainly at high risk on May 18. In the Gardener, NY, frost event noted earlier, we found large cankers in several old trees that may have contributed inoculum for that devastating frost-trauma event. Also, Decker and Daemen (1999) noted that presence of over-wintering cankers appeared to be critical for frost-induced blight infections on pear trees in Europe.
In conclusion, I suspect that risks of frost-induced blight are relatively low this year because the frost occurred after bloom (in most orchards), weather was dry during the preceding days, and inoculum levels therefore should have been too low to allow frost-induced infections. However, growers should be alert for symptoms of fire blight next week (May 29-June 3). That’s the good news. The bad news is that frost-induced fire blight, when it appears, may be so severe that it will be impossible to remove infected shoots via pruning. If that occurs, I see only two options. On young trees (<8 years old), it may be possible to restart badly infected trees by cutting off the entire top of the trees below the last branch if such pruning is done before the bacteria can be translocated down into the lower trunk or to the rootstock (if it is a blight susceptible rootstock). Detail pruning to salvage tree structure will likely be a waste of time on young trees severely affected by frost-induced trauma blight. On older trees that have settled down, it may be best to walk away and allow blight to take its course if there are scores of infections per tree, then salvage what remains during winter pruning next year. In both cases, if severe infection is limited to just a small number of trees, then removing those trees is the best option.
More suggestions on using the NEWA fire blight model: For those who wish to check their local conditions around the time of the frost event, the following may be helpful.
- On NEWA, go to the “Crop and IPM tools”, then scroll down to “Fire blight”.
- Enter a NEWA station location.
- When the fire blight page opens, if you wish to see the EIP levels prior to the frost, enter a blossom date. The actual date is not important at this point, so just enter April 30.
- For “Date of interest”, enter May 17 and the table will show EIP levels for May 15-22.
- If you wish to see effects of strep sprays applied prior to the frost, scroll down past the infection and wetness tables and enter a spray date for strep, then scroll back up to the EIP table to see how strep applied on that date would have affected EIP levels. For locations that I tested, it appears that strep applied any time after May 12 should have kept the EIP below 100 right up through the frost event on May 18. (Not sure how relevant that is?)
- If you wish to see WHEN frost-induced trauma blight might appear in your orchard, clear the date for first blossom opened, click on May 18 as the date of interest (while leaving first blossom date blank), then scroll down to “Scouting for shoot blight associated with a trauma event” and enter May 18 (the date of the frost). That will trigger a predicted date for appearance of trauma blight symptoms. For some reason, the option for entering a trauma blight date disappears if you enter a date for first bloom.
One final comment: Frost events can also trigger shoot blight caused by Pseudomonas syringae, a bacterial pathogen that likes cool damp weather. On pears, this causes pear blast disease, which looks very much like fire blight. A similar blast disease caused by P. syringae has very occasionally been seen on apples following frost injury. I suspect that the dry weather prior to the May 18 frost and the lack of flowers on most trees at the time of the frost did not favor development of Pseudomonas infections this year.
Nevertheless, it is important to distinguish pear blast from fire blight. Pear blast will NOT spread beyond the current season growth and will not produce ooze droplets on infected tissue like fire blight does. Removal by pruning is not important for managing pear blast, whereas removing fire blight strikes via pruning is advised unless the number of infections is so large as to make it unfeasible.
Literature cited:
Deckers, T. and Daemen, E. 1999. Determining factors reducing the host susceptibility of fruit trees for fire blight infections. In: Proc. 8th International Workshop on Fire Blight, Momol and Saygili (eds.) Acta Horticulturae 489:483-489
