Sustainable Management of Fire Blight in Apples using Plant Growth Regulators and Plant Defense Activators

Year 1

Disease Management. At Cornell AgriTech, in the mature semi-dwarf block, the untreated trees experienced 88.1 ± 3.3 and 54.7 ± 5.4 % incidence of blossom and shoot blight respectively. All treatments experienced significantly less blossom and shoot blight (p < 0.0001, Table 4, Figure 1). Antibiotic programs performed the best with a mean of less than 5.7 ± 3.7 or 1.0 ± 1.0 % blossom and shoot blight respectively. The experimental treatments frequently provided similar (non-significantly different) levels of disease control as compared to antibiotic programs. Pre-bloom PhCa applications at both rates and timings experienced less than 20% or 10% blossom or shoot blight incidence respectively.

At University of Vermont in the mature orchards, treatments frequently had a significant effect on blossom blight incidence, although effects were not as great as at the Cornell AgriTech orchard (Figure 2, Table 4). The untreated trees experienced 64.0 ± 15.4 and 18.0 ± 10.0 (‘Crimson Crisp’) or 69.0 ± 10.0 and 13.3 ± 7.8 (‘Topaz’) percent incidence of blossom and shoot blight respectively. For ‘Crimson Crisp,’ only antibiotic programs experienced significantly less blossom blight than untreated trees, with a mean of 10.9 ± 5.2 % (p < 0.0001, Figure 2). For Topaz, all experimental programs had less blossom blight than the untreated trees (p < 0.0001, Figure 2). Antibiotics performed the best (19.2 ± 9.7 %), followed by Regalia (26.2 ± 5.6 %), and then by the pink PhCa programs, ranging from 39.5 to 44.5 % incidence. Shoot blight pressure was not adequate for either cultivar to draw conclusions about experimental treatments.

Horticultural Impact. At Cornell AgriTech in the mature semi-dwarf block, minimal differences were observed between treatments with regard to horticultural parameters (Table 4). There was a significant effect on shoot length at harvest (p < 0.0001, Figure 3). Post-bloom applications of PhCa reduced shoot length, as expected; tight cluster application of PhCa at the higher rate also reduced shoot length as compared to the untreated trees. Trunk size (p = 0.2851) and fruit size (p = 0.0730) were not impacted by treatments. Fruit number was influenced by treatment (p = 0.0096), with the tight cluster application of PhCa at the higher rate having the higher fruit number than the streptomycin application at bloom; however, neither were different than untreated trees. 

At University of Vermont in the mature semi-dwarf block, similarly, experimental treatments had a minimal impact on horticultural parameters (Tables 5 and 6). Shoot length at harvest was impacted for both ‘Crimson Crisp’ (p = 0.0023) and ‘Topaz’ (p = 0.0019). For ‘Crimson Crisp,’ the lower rate of PhCa at pink and the PhCa+Actigard treatments had shorter shoots than the untreated trees. For ‘Topaz’ the lower rate of PhCa at pink had longer shoots than the Regalia treatment, but neither were different than the untreated control. More importantly, shoot growth in this block (both varieties) was minimal due to low vigor (<20 cm of growth), and this data is likely not an accurate representation of whether that shoot length was compromised by treatments.   

At Cornell AgriTech in the young high-density planting, experimental treatments did not have an effect on any of the horticultural parameters (Table 7) evaluated including shoot length (p = 0.4071), trunk diameter (p = 0.2337), fruit number (p = 0.1207), and fruit size (p = 0.4217).

At University of Vermont in the young high-density planting, experimental treatments also did not have an effect on any of the horticultural parameters evaluated (Table 8), including shoot length (p = 0.3387), trunk diameter (p = 0.8405), fruit number (p = 0.1596), and fruit size (p = 0.5535).

Year 2

In year two, we observed very similar results compared to the first year’s evaluation.

Disease incidence. In the mature orchard at Cornell AgriTech, disease incidence in untreated control trees was 82.0 ± 6.4 and 42.6 ± 5.9 % incidence for blossom and shoot blight, respectively, indicating disease incidence was sufficient for reliable evaluation of treatments. Both blossom and shoot blight were significantly affected by treatments (Table 9). Blossom and shoot blight were both significantly lower for all treatments than Untreated trees (p < 0.0001). No difference was detected between other treatments. Disease incidence was always less than 30% and 13% for blossom and shoot blight respectively. In addition, no differences were detected between treatments for any of the horticultural parameters evaluated.

In the mature ‘Crimson Crisp’ and ‘Topaz’ orchards at University of Vermont, blossom blight incidence in the Untreated treatment was 84.4 ± 3.4 % and 60 ± 7.5 %, respectively, again indicating sufficient disease pressure for evaluation of treatments. In both orchards, disease incidence was significantly effected by treatment (Tables 10 and 11). Strep+PhCaPB was the most effective at reducing blossom blight in both orchards with 52.5 ±8.4 and 9.5 ± 4.0 % incidence, in respective orchards. In the ‘Crimson Crisp’ orchard, only Strep+PhCaPB and PhCa+Actigard significantly reduced blossom blight compared to the Untreated treatment (p < 0.0001). In the ‘Topaz’ orchard, all the experimental fire blight programs successfully reduced blossom blight incidence (p = 0.0006). Shoot blight incidence was again very low in the Untreated treatment with 21.0 ± 4.3 % and 1.4 ± 1.1 % incidence in the ‘Crimson Crisp’ and ‘Topaz’ orchards, respectively. None of the experimental treatments resulted in significantly different shoot blight control than the Untreated trees (p = 0.2291 and 0.7334).

Horticultural Impact. In the mature ‘Empire’ orchard at Cornell AgriTech as well as the mature ‘Crimson Crisp’ and ‘Topaz’ orchards at University of Vermont, no significant effects were detected for any of the horticultural parameters that were evaluated (Tables 9-11, p > 0.05).

In the young, high-density plantings, minimal impact was observed on horticultural parameters at both Cornell AgriTech (Table 12) and at the University of Vermont (Table 13). In the ‘NY-1’ orchard at Cornell Agritech, shoot length was impacted in both July (p = 0.027) and September (p = 0.0012) measurements. In both cases, the PhCa Trickle program significantly reduced shoot length as compared to Regalia and PhCa 6oz in July and the Untreated trees at harvest. In the ‘Macoun’ orchard at University of Vermont, horticultural parameters were never impacted by treatments (p > 0.05).

Year 3.

In year three, the pandemic travel restrictions in early 2021 forced us to conduct both trials at Cornell AgriTech as opposed to a third year at UVM. With the NCE, we plan to conduct a third year at UVM this spring pending travel restrictions if any.

Disease incidence.

Across both sites, the weather was cool and not conducive for the development of fire blight until the end of petal fall and the weeks following. In the tall spindle ‘Idared’ orchard, the incidence blossom and shoot blight in untreated control trees was 68.0 ± 3.1 and 55.3 ± 3.4 %, respectively, suggesting that disease incidence was sufficient for reliable evaluation of treatments. The incidence of blossom and shoot blight for all experimental PGR treatments was significantly lower than that of the ‘untreated trees’ (Table 14; P < 0.0001). There were also no less effective than the streptomycin/prohexadione Ca industry standard treatment, and there were no differences in fire blight incidence among any of the experimental PGR treatments (Table 14; P > 0.0001); For these experimental treatments, the incidence of blossom and shoot blight was always less than 10% and 5% for respectively. Additionally, there no differences among any of the treatments for any of the horticultural parameters evaluated.

In the high-density ‘Evercrisp’ orchard, the incidence blossom and shoot blight in untreated control trees was 35.7 ± 2.9 and 14.6 ± 0.3 %, respectively. The lower disease incidence in the high-density block, likely resulted from increased air circulation due as trees in this planting system have less dense canopies. Overall, the levels of disease were sufficient for reliable evaluation of treatments. The incidence of blossom and shoot blight for all proposed treatments was significantly lower than that of the ‘untreated trees’ (P < 0.0001) and were not different from the streptomycin/prohexadione Ca industry standard treatment (Table 15; P > 0.0001). Also, there were no differences in fire blight incidence among any of the other treatments. For the experimental PGR treatments, the incidence of blossom and shoot blight was always less than 3% and 2% for respectively. Additionally, there no differences among any of the treatments for any of the horticultural parameters evaluated.

Horticultural Impact. In the tall spindle ‘Idared’ orchard there were no significant effects among treatments for any of the horticultural parameters evaluated (Table 14; P > 0.05). Hence, there was no impact of the experimental growth regulator treatments on the development of the trees by mid-September. By contrast, there industry standard streptomycin/prohexadione Ca treatment had a slight impact on the development of shoots in the high density planting by mid-September compared to the other treatments (Table 15; P < 0.05). Aside from shoot growth, trunk diameter and yield were unaffected in the high density planting.

Year 4 NCE UVM site.

In year three, the pandemic travel restrictions in early 2021 forced us to conduct both trials at Cornell AgriTech as opposed to a third year at UVM. We requested and received an NCE to conduct a final year at UVM. In the early spring and summer of 2022, we set up another trial on ‘Crimson Crisp’ to evaluate blossom and shoot blight in May of 2022. Given the termination of the award in June we were unable to collect data on the physiology and harvest of the crop in September.

Disease incidence.

At UVM, the weather was exceptionally hot and wet at the end of the bloom period and was quite conducive for the development of fire blight. This was especially the case at the end of petal fall and the weeks of shoot development. In this mature ‘CrimsonCrisp’ orchard, the incidence blossom and shoot blight in untreated control trees was 73.0 ± 7.0 and 45.5 ± 10.9 %, respectively, suggesting that disease incidence was sufficient for reliable evaluation of treatments. The incidence of blossom and shoot blight for all experimental PGR treatments was significantly lower than that of the ‘untreated trees’ (Table 15; P < 0.0001). There were slightly less effective than the streptomycin/prohexadione Ca industry standard treatment, but there were no differences in fire blight incidence among any of the experimental PGR treatments (Table 15; P > 0.0001); For these experimental PGR treatments, the incidence of blossom and shoot blight was nearly 50% and 25% for respectively. Additionally, there no differences among any of the treatments for any of the horticultural parameters evaluated.

Horticultural Impact. In the mature ‘CrimsonCrisp’ orchard there were no significant effects among treatments for any of the horticultural parameters evaluated (Table 15; P > 0.05). Hence, there was no impact of the experimental growth regulator treatments on the development of the trees by late June, when the award ended. Even the industry standard streptomycin/prohexadione Ca treatment no impact on the development of shoots compared to the other treatments (Table 15; P < 0.05). Aside from shoot growth, trunk diameter and yield in June were unaffected in the high-density planting.

Figure 7. Disease incidence blossom blight (left), shoot blight (right), and shoot growth (bottom) in mature high-density ‘CrimsonCrisp’ block at UVM in 2022.

Table 15. Effect of fire blight management programs on disease and horticultural parameters at University of Vermont in a mature tall-spindle ‘Crimson Crisp’ orchard 2022

Discussion

All experimental treatments the established blocks at Cornell AgriTech show promise for managing fire blight even after three years of testing. The disease management programs provided both blossom and shoot blight control comparable to the streptomycin/prohexadione Ca industry standard program and may be appropriate for commercial use. However, results from the UVM orchard for both varieties in the mature, tall-spindle planting were less promising. None of them provided adequate disease control for blossom or shoot blight. However, several factors may be influencing results. This block was historically managed as organic certified, and the soil is extremely sandy. Results from the 2019 season indicated that adequate tree vigor may be important for the PGR PhCa and/or plant defense inducers to be effective for fire blight management. In the second season of this trial, the fertilization program included increased nitrogen to maintain normal tree nutrition and tree growth. As a result, we saw slightly more shoot growth in 2020 than in 2019, and slightly better disease control. In the 2021 season, we implemented the trials in a young EverCrisp block with a higher nitrogen program and observed a stronger response from the PGR and plant defense inducers resulting in excellent fire blight control. Similarly, a stronger response was observed in the ‘Topaz’ orchard as compared to ‘Crimson Crisp’ orchard in 2019 and 2020 for many of the experimental treatments. Interestingly, the 2022 season was  warmer with adequate rainfall, the trees were more active, and we observed the best response and affect from the PGR programs in all of the years of our trials at UVM

Several factors may contribute to this difference including 1. ‘Topaz’ blooms slightly earlier and 2. the materials evaluated require time to activate host defense response. As a result, it is possible that specific timing of applications to phenology in advance enough may be important for providing disease control. Another explanation may be the differential host-resistance for specific varieties. Our final season evaluating these treatments at UVM on ‘CrimsonCrisp’ further suggest that the warm weather, fertilizer, and adequate rainfall have influenced the efficacy of these PGR and plant defense inducer treatments for fire blight disease management. The results from these four seasons of trials suggest tat there may be variability in the efficacy of PhCa and plant defense inducers for management of fire blight, regarding region, cultivar, and/or tree vigor. However, as we become more familiar using these materials in the specific locations and cultivars over the four years, we were able to improve effectiveness of the practices. It suggests that grower should consider they own block and cultivar nuances while implementing these practices. These results are likely to be valuable for informing more nuanced recommendations for using PhCa and plant defense inducers to manage fire blight. 

In all four years and locations, plant growth regulator and plant defense inducer programs had minimal effects on horticultural parameters in mature tall-spindle plantings and they did not affect any of the measured parameters in the young high-density plantings. The main exception we observed was for the post-bloom applications of PhCa, which were only in the Streptomycin/prohexadione Ca industry standard program. These programs, in some years, locations, and varieties, slightly reduced shoot vigor. Where tree growth was impacted, it was minimal, typically on the order of 10 cm or less. This amount of growth reduction is likely not horticulturally significant, especially in a mature block of larger trees. There is more concern about shoot growth reduction for young, high-density trees, which are pushed to fill canopy space in the first 2-3 years of establishment. In the three years of this work, young high-density blocks were not influenced by the experimental treatments.  

Overall, our results indicate that pre-bloom PhCa programs alone and in combination with plant defense inducer products (Actigard and a phosphite fungicide) may be effective fire blight management programs for mature tall-spindle trees, and without negatively impacting tree growth or productivity. These results are consistent with our results from other earlier trials in a Gala tall-spindle planting in Cornell AgriTech Research Orchards. These prohexadione-calcium programs offer promising alternatives to fire blight management using antibiotics.