New England Cider Apple Program: Optimizing Production for High-Value Markets

Project activity was affected in 2020 by the COVID-19 situation, which delayed some work and prevented other efforts. Early-season (March-June) travel and orchard access was limited for most cooperators; a COVID outbreak on one participating farm in Vermont prevented access at harvest to assess fruit and juice characteristics. Activity in 2021 and 2022 resumed as planned, although some activities were restricted  due to lack of in-person meetings, growers dropping out of cider production, and loss of project personnel.

Cider cultivar observations (Maine, Massachusetts, Vermont).

Cider apple cultivar data collection instruments were trialed in 2019 by Bradshaw and Garofalo. Conclusions from that season were limited, and during winter 2019-2020 and in successive growing seasons, data were compared with those from grower cooperators to validate individual observations.
In 2020 and 2021, twenty-five cider apple cultivars were evaluated across ten farms. Observations were limited to sometimes one-off farm visits due to COVID-related travel restrictions, so not all cultivars were observed for multiple years or across the full season.
Some general trends observed include:

1. Most European cider apple cultivars show substantial biennial bearing habit, as expected and reported in literature and grower observations.
2. Many of the European cider apple cultivars, e.g., Dabinett, Yarlington Mill, Michelin, and others, have a pronounced susceptibility to fire blight infection.
3. North American heirloom cultivars are variable in their annual bearing habit, but as a group have substantially less issue with biennialism than European cider varieties.
4. One class of cultivars that may be useful to growers includes Scab-resistant cultivars that may be easier to grow and show, in the lab, intermediate juice chemistry characteristics between traditional dessert cultivars and high-phenolic European and similar cider cultivars.

Scab-resistant apple cultivars may provide a good balance between moderately high phenolics desirable in cider making and disease-resistance.

Mechanical thinning research (Maine)

Spur pruning trials at Highmoor farm in Monouth, ME.

The string thinning experiment was put on hold due to the Covid-19 pandemic and increased fire blight in the intended research orchard.  Instead, observations on string thinning effectiveness were done at a commercial orchard in Greene, ME with Gala, a variety that may have larger fruit size from prebloom thinning.  The grower’s method was modified based on results from the previous year’s experiment.  In 2019, we learned that the string thinner overthinned outer branches and did not thin the inner canopy branches.  To counteract this imbalance, I suggested to the grower that he focus on thinning the outer branches only and adjust the speed of the string revolutions so that overthinning no longer occurred.  Follow-up observations indicated that trees were not overthinned, but also that they may not have been thinned enough.  Due to physical distancing requirements, I was not able to properly mark the rows that were thinned, so detailed measurements were not made.

The spur pruning (artificial spur extinction) experiments in the Honeycrisp rootstock trial continued in 2021 as planned.  In April, two rows of trees (35 trees per row) were spur pruned and one row was not.  We spur pruned the same trees that were spur pruned in 2019 and 2020. Control trees remained without spur pruning, as a way to document repeated spur pruning on bloom and yield, but also on the amount of time required to continue the practice. Limb renewal plus simplifying (removing side branches but keeping spurs) took 1 to 2 minutes for a dwarf-sized trees. Return bloom counts were complicated by the need to do spur pruning three weeks before bloom, so bloom count data may not fully represent return bloom. Bloom and fruit counts were done on all 12 rootstocks (Table 2). 

Most of the trees were in an ‘off’ year of the biennial cycle, so spur pruning was not as severe as in previous years. In this the third year of consecutive pruning, bloom and crop load were greater with spur pruning, an important attribute in an off year (Table 2). The lighter crop load resulted in larger fruit size in the control trees, so the impact on yield per tree was minimized. Biennial bearing index (change in yield from 2020 to 2021) was not significantly greater with the control treatment. Yield per acre and cumulative yield did not statistically differ between the two pruning methods in 2021, but spur pruning moderated the up and down cycling in yield (Figure 1).  From a practical standpoint, spur pruned trees produced over 100 more bushels than trees not spur pruned.

Interactions occurred between spur pruning and rootstock, so data from four selected rootstocks is shown in Table 2. Spur pruning moderated year-to-year fluctuations in crop load in one rootstock, G.214, but not the other three. Spur pruning moderated fluctuations in yield in three of the rootstocks, but not in the most vigorous and least biennial one, V.1.

Figure 2. Honeycrisp crop load (number of fruit per tree) and yield per tree in spur pruned and without spur pruning (control) in four rootstocks. Spur pruning began in 2019. 

We began additional spur and limb pruning experiments in three Honeycrisp orchards in 2021. In the Turner orchard, five trees were spur pruned and another five were not. Trees were spur pruned to singulate spur clusters and to remove flower buds where there were several occurring in close proximity to each other.  All trees were previously pruned by the grower. Observations in two entire rows of trees indicated that 22% (12 out of 54 trees) of the trees had no bloom in 2021. These trees were tagged for future bloom counts. In Orchards 94 and 7-202, located at the experiment station, we added a limb removal treatment to compare with spur pruning. These trees were previously lightly pruned by the farm manager.

Spur pruning reduced the number of flower clusters in two of the orchards (Table 3). The effect was minimal in the Turner orchard which had fewer flowers than the other orchards. Limb removal was not as time consuming as spur pruning and reduced the number of flower clusters, but not enough to substantially reduce crop load.  Pruning treatments did not influence fruit quality.  

1. Return bloom research (Vermont)

Hedge pruning trials at University of Vermont Horticulture Research and Education Center (South Burlington, VT) and Sunrise Orchards (Cornwall, VT).

Three mechanical hedging timings were compared for their effects on return bloom in cider cultivars trained to tall spindle. Treatments consisted of 1) Normal winter dormant pruning with hand tools as a control. 2) Mechanical winter dormant pruning with hedger. 3) Mechanical pruning at pink (prebloom) bud stage with hedger. 4) Mechanical pruning at 12-14 leaf stage (mid-June) with hedger. Treatments were applied in a randomized complete block design, with six single-tree replications per treatment.

Replicated field trials were done at two orchards in Chittenden and Addison County, Vermont. The first site located at the UVM Horticulture Research and Education Center in South Burlington, VT.  Productive, 8 year old, ‘Empire’ and ‘McIntosh’ trees grafted onto 'Budagovsky 9’ (BUD 9) rootstock spaced at 0.9m · 4.5m apart in a Windsor Adams loamy sand with supplemental irrigation were selected for the trial. The second location, a commercial orchard in Cornwall, VT with hard cider cultivars ‘Somerset Redstreak’ and ‘Harry Masters Jersey’ grafted on NIC29 ® rootstock established in 2016 planted at 0.9m · 4 m spacing in Vergennes clay soil with supplemental irrigation. At both sites, orchard floor management consisted of semi frequent mowing of the interrows with a 1-m herbicide strip maintained in the intrarow. Trees were irrigated at the grower’s discretion. Each planting followed standard commercial practices for pest and fertilization management. 

Hedging was performed using a mechanical hedge trimmer (STHL model KM 56 RC-E with HL-KM attachment, STIHL Inc. Virginia Beach, VA). Trees were trimmed to a fruiting wall measuring two feet across the row using a measured guide attached to the trimmer. Hedging performed during the growing season was completed when no rain was forecast for two days following the procedure to limit potential for firelight infection. 

‘Somerset Redstreak’ was recorded at full bloom on May 21 in 2019 and 2020.  Fruit was harvested according to growers schedule and recommendation on 16 Sept 2019 and 8 Sept 2020. ‘Harry Masters Jersey’ full bloom was recorded on 27 May 2019 and 29 May 2020. Harvested 25 Sept 2020, 2019 harvest date unrecorded. ‘McIntosh’ came into full bloom 23 May 2019 and 21 May 2020, harvested on 19 Sept 2019, and 24 Sept 2020. ‘Empire’ full bloom 24 May 2019, 21 May 2020 with harvest on 26 Sept 2019, and 28 Sept 2020.  

Data Collection

At full bloom, for each treatment-replicate, the total number of flower clusters on each tree was counted and recorded. Each fall vegetative growth parameters: tree height and spread (m), trunk circumference (cm), and the length of five terminal branches per tree were measured. At harvest total crop yield was measured (kg tree-1) and number of fruit per tree recorded. The number of recently dropped fruits were recorded separately and assumed to be of average fruit weight. A randomly selected sample of fruit (5) per treatment-replicate (tree) was collected from harvested fruit and assessed for fruit size, scored for percent green background color to red foreground color, general defects, and USDA grade distribution (Bradshaw et al., 2018). After external evaluation, internal fruit qualities such as fruit firmness and ripeness were assessed. Fruit firmness was measured using a 11-mm probe penetrometer (Wagner, Greenwich,CT) and ripeness assigned using the starch iodine index (Blanpied; Silsby 1992-07 #125). Fruit samples were then analyzed for juice quality parameters including pH, titratable acidity, total phenolics, and soluble solids using standard protocols. 

Data analysis

Data were subject to analysis of variance (ANOVA) procedures by hedging treatment separately for each orchard location and year (SAS Institute Inc., 2002-2010). If overall variances were found at α=0.05, post-hoc multiple comparisons were made using Tukey’s adjustment.

Hedging treatments applied in 2019 were expected to affect the following season’s flowering.  In 2020, the dormant hand pruned treatment on ‘McIntosh’ had an average return bloom of 238 flower clusters per tree, 83% of the prior year’s total. Hedging treatments on ‘McIntosh’ in 2019 had a 47-54% reduction in the number of returning flower clusters from the prior year (Table 4). This could be because there was less canopy volume and foliage to support flower development. There were no differences in bloom attributable to hedging treatment in 2021. In 2019 a trend toward reduced yield per tree and yield efficiency (yield per TCA) from hedging was observed for almost all cultivars, the exception being ‘McIntosh’. Dormant hand pruned ‘McIntosh’ had an average 6 kg yield increase over any hedging treatment in 2020.  ‘Empire’ hedged at pink bud stage had higher crop yield than dormant hand pruned trees in 2020. After the second year of hedging ‘Somerset Redstreak’ showed an increase in cumulative yield for all hedging times. The cropping of ‘Somerset Redstreak’ in 2020 shows the potential for annual bearing tendencies for that cultivar. Previous studies on hedging ‘Empire’ suggests that hedging increases cumulative yield over a ten-year period (Ferree and Rhodus, 1993). Two years of data presented here are currently unable to fully support that statement, but hedging appears to be a promising management tool for both ‘Empire’ and ‘Somerset Redstreak’. ‘Harry Master Jersey’ exhibited biennial tendencies with very few trees flowering in 2020.  This confirms the tendency for ‘Harry Master Jersey’ to be biennial and that following one-year of hedging, return bloom was not stimulated. More data are necessary to confirm this trend.

Hedging in 2019 narrowed the spread of the trees, reducing the canopy volume for each cultivar (Table 5). Hand-pruned trees were nearly twice the size of trees pruned mid-June. Hand pruned trees remained larger in 2020, but ‘McIntosh’ and ‘Somerset Redstreak’ did not show a difference among treatments. Summer hedging at pink stimulated shoot growth for ‘Somerset Redstreak’ and ‘McIntosh’ leading to a wider and denser canopy. No TCSA differences were observed in 2019 or in 2020 between trees hand pruned or hedged, across all four-cultivars surveyed.

There were no differences in juice quality for soluble solids, pH, titratable acidity, or total phenolic among treatments.  Juice chemistry (Table 6) was within a normal range for all cultivars (Alexander et al., 2016; Bradshaw et al., 2018). This shows that canopy management done throughout the year and altering tree structure via hedging does not negatively affect juice quality. These results provide apple growers and hard cider producers with a better understanding of how different crop load and canopy management strategies influence juice quality at harvest. Fruit quality parameters (Table 7) for red color, firmness, and starch index rating remained unaffected by treatment. Although summer hedging opened up the canopy to allow potentially more light penetration into the tree, there were no increases in fruit color observed on hedged trees.  There were no differences in fruit firmness and starch indexes at α error of 0.05. This did not agree with previous work which showed increases in fruit color, softer fruit, and higher starch indices on summer pruned ‘McIntosh’ trees (Schupp, 1992).

Hedging during the summer caused tissue damage to tree limbs and shoots, leaving a splintered ‘broomstick’ effect on the end of trimmed branches. Pruning trees mid-season can carry an increased risk of fire blight infection. Fire blight caused by the bacterium Erwinia amylovora is a destructive disease that causes dieback of blossoms, shoots, limbs and under ideal conditions can kill the tree. Hedging at pink and mid-June causes wounds that fire blight bacterium can enter. Infected trees can develop lesions that ooze orange bacterium filled liquid that is easily spread in moist, warm weather, by splashing rain, dew, wind and insects. The use of hedging equipment can also spread disease if not properly sanitized. Damaged branches that have dead tissue also have the potential to host a range of fungal diseases, such as black rot (Botryosphaeria obtusa) that can infect fruit and form cankers. Disease management for ‘McIntosh’ and ‘Empire’ is well understood, but the fire blight and disease susceptibility of ‘Harry Master Jersey’ and “Somerset Redstreak’ grown in the northeastern U.S.A. is less well-established. No fire blight damage was seen in this study, likely in part due to proper sanitation and timing hedging treatments around weather conditions. Growers would benefit from a robust disease assessment on damage caused by hedging and the incidence of disease.

Results of this study suggest that summer canopy management does not alter apple juice quality. ‘Harry Masters Jersey’ showed a tendency for biennial bearing, and summer hedging was unable to stimulate return bloom. ‘Empire’ and ‘Somerset Redstreak’ may both benefit from hedging showing signs of increased yields and annual bearing. Future studies should continue to record flowering and yield of ‘Somerset Redstreak’ and ‘Harry Masters Jersey’ to establish a biennial bearing index. Based on the two years of data presented, ‘McIntosh’ trained to tall spindle may not be suitable for hedging due to decreased yields and flower return. Both cider cultivars would benefit from a specific crop load management study that hand thins trees to specific fruiting densities based on TCSA. More research-based information is needed to understand the flowering and cropping of specialty cultivars to inform growers on how to maintain consistent annual production. 

 This material was presented at in June 2021 at XII International Symposium on Integrating Canopy, Rootstock and Environmental Physiology in Orchard Systems.

Foster, J. Kingsley-Richards, S.L., and Bradshaw, T. 2022. Effect of Summer Hedging on Return Bloom, Yield, Tree Growth, and Juice Quality of Apples Grown for Cider. Acta Hortic. 1346, 439-446. DOI: 10.17660/ActaHortic.2022.1346.56

1. Plant growth regulator trials at University of Vermont Horticulture Research and Education Center (South Burlington, VT) and Sunrise Orchards (Cornwall, VT).

For commonly grown dessert apple cultivars chemical thinning or the removal of some fruit each season helps maintain fruit size, quality, and annual bearing characteristics. Chemical thinning is often achieved with applications of carbaryl at petal fall alone or in combination with other plant growth regulators (PGRs). This traditional thinning program used for dessert fruit does not adequately thin European-origin cider apples resulting in insufficient return bloom or inconsistent cropping from year to year. On dessert apple cultivars with biennial bearing tendencies, midsummer applications of PGRs are used to enhance fruit bud development for the following year. In  2019, experiments were conducted in two apple orchards in Vermont, U.S.A with the primary objective to evaluate  the effects of naphthaleneacetic acid (NAA) and ethephon alone and in combination with carbaryl on return bloom, crop yield, and fruit and juice quality. During the two years of the study, cultivars ‘Harry Masters Jersey’ and ‘Kingston Black’ both demonstrated strong biennial production habits producing few flowers and fruit in 2020. Ethephon applications alone and in combination with carbaryl showed advanced ripening and fruit softening in ‘Somerset Redstreak’ during the year of treatment. ‘Kingston Black’ had increased fruit softening with ethephon only applications. Growth regulator treatments did not have a consistent effect on juice quality between cultivars. During the treatment year, 2019, all Ethephon treated ‘Somerset Redstreak’ had a higher pH and juice from trees treated with Ethephon and carbaryl had a lower titratable acidity.  ‘Kingston Black’ juice was unaffected by PGR applications.

This material was presented at in June 2021 at XII International Symposium on Integrating Canopy, Rootstock and Environmental Physiology in Orchard.

Foster, J. Kingsley-Richards, S.L., and Bradshaw, T. 2022. Summer Applications of NAA and Ethephon Show Little Effect on Return Bloom, Yield, Tree Growth, and Juice Quality of Cider Apple Cultivars. Acta Hortic. 1346, 511-518. DOI: 10.17660/ActaHortic.2022.1346.65