Materials and Methods

‘McIntosh’ and ‘Honeycrisp’ fruit were grown, harvested, and stored in three regions with similar growing conditions: Monmouth, ME (lat. 44°13′51″ N, long. 70°4′5″ W), Lake City, MN (lat. 44°51′30″ N, long. 93°39′41″ W), and Norfolk County, ON (lat. 42°52′44″ N, long. 80°15′22.6″ W). The same orchard in each location was used in both 2016 and 2017. Although only one orchard was sampled in each region, each represents different soil and weather conditions of the Midwest and Northeastern United States. In ME, trees were grafted to ‘M.26’ and ‘Geneva 16’ rootstocks and were planted in 2002. ‘Honeycrisp’/‘M.26’ trees in ON were planted in 1998. In MN, trees grafted to ‘B.9’ were planted circa 1997. In 2016, fruit from ‘Roger’s Red McIntosh’/‘M.7 EMLA’ in ME, ‘Summerland McIntosh’/‘M.7’ in ON, and ‘Ruby Mac’ in MN were included in the study.

Date of harvest was based on preliminary maturity estimates of a few fruit using SPI. In 2016, at each location, fruit were harvested randomly, without regard to color, size, or canopy position, three times at weekly intervals beginning 8, 13, and 14 Sept. in MN, ME, and ON, respectively, for ‘McIntosh’; and 31 Aug., 14 and 15 Sept. for MN, ON, and ME, respectively, for ‘Honeycrisp’, except in ON where fruit were harvested twice. Replications each consisted of at least three trees to minimize individual tree effects and 45 to 70 fruit per replication in ME; 40 to 75 ‘Honeycrisp’ and 40 to 120 ‘McIntosh’ fruit per replication in MN; and 85 ‘Honeycrisp’ and 100 ‘McIntosh’ fruit per replication in ON. In 2017, three replications each with 60 fruit were randomly harvested three times at weekly intervals beginning 18 Sept. in MN. In ME, 55 to 80 apples were harvested from each of two replications at weekly intervals beginning 15 Sept. In ON, one replication of 250 fruit from several trees were harvested at weekly intervals beginning 7 Sept. ‘McIntosh’ apples were not studied in 2017.

Apples were numbered with a fine-tip permanent ink marker on the peel at the interface of sun and shade exposure and near the stem end to identify fruit and later match I_AD at harvest with post-storage disorders and fruit quality, but also to indicate where to sample peel for chlorophyll analysis. The same day or one day after harvest, I_AD was measured on two sides of each apple on an area below the ink mark and on the opposite side of the fruit, avoiding any areas with sunburn. If I_AD readings on opposite sides of an apple substantially varied from one another, an additional reading was made to confirm or reject the disparity. Values that indicated potential sunbleaching were rejected. All measurements were made with a Delta Absorbance Meter (Sinteleia, Bologna, Italy). Fruit were grouped in increments of 0.2 I_AD units for fruit quality measurements. Fruit weight of each apple and the amount of fruit surface area (2017 only) with blush were measured on each apple. Five to ten fruit from each I_AD group were used to measure SPI, soluble solids content (SSC), and firmness on two peeled sides of each fruit. Some I_AD groups had insufficient fruit for both maturity and storage, so harvest measurements were not made for these groups. Starch staining with iodine was measured by dipping or spraying each cross-sectioned apple in or with a potassium-iodine solution and using a visual rating where 1 = all starch remaining and 8 = no starch (Blanpied and Silsby, 1992). Flesh firmness was measured using a drill press-mounted penetrometer (McCormick Fruit Tester model FT 327, Alfonsine, Italy) in MN; an electronic texture analyzer (Guss Manufacturing, Strand, South Africa) in ON; and an EPT-1 (Lake City Technical Products, Kelowna, BC, Canada) in ME; all equipped with an 11-mm diameter plunger. Soluble solids concentration was measured using a hand-held temperature-compensated refractometer [Atago model PAL-1 3810A in ME and MN; and an Atago PR-32 in ON (Tokyo, Japan)] from a pooled sample in ME and ON and from individual fruit in MN. Juice was expressed during pressure testing.

Internal ethylene concentration (IEC) was measured at harvest and after storage on fruit harvested in ON. A 3-mL gas sample was withdrawn from the core of each of five fruit within an I_AD group using a syringe and then injected into an Agilent 7820A gas chromatograph (Agilent Technologies Canada Inc., Mississauga, Ontario, Canada) equipped with a 0.25-mL sample loop, flame ionization detector, and a 25-m × 0.53-mm CarboBOND capillary column (Agilent Technologies Canada Inc.). The injector, column, and detector temperatures were 150, 80, and 250 °C, respectively. High-grade helium was used as the carrier gas, with a typical run time of 1.5 min.

In all three locations, fruit were stored at 0.5 °C in air for 2 months in 2016 and 4 months in 2017. The longer duration was tested in 2017 to allow for full development of storage disorders. Following storage, incidences of soft scald, soggy breakdown, bitter pit, lenticel blotch, lenticel breakdown, senescent breakdown, diffuse flesh browning, and core browning were counted.

Results

I_AD at harvest.

In both years, ‘Honeycrisp’ fruit peel I_AD varied among harvest dates and location with a significant interaction between the two factors (Table 1). In 2016, I_AD was greater in MN than in ME and ON during each harvest. It also decreased with later harvest date in MN and ME, but not in ON where the two harvest dates were only 5 d apart. In 2017, I_AD decreased with later harvest date in all three locations, but the decrease in MN was not significant from harvest 2 to harvest 3, which is similar to what occurred in 2016 in this location. In 2017, I_AD was consistently higher in ME than in ON and higher than in MN during the first harvest.

Table 1.

Fruit peel I_AD at three harvest times in ‘McIntosh’ and ‘Honeycrisp’ apples grown in three regions.

View Table

A wide range in I_AD occurred within a harvest date in ‘Honeycrisp’ in both years (Fig. 1). In MN, I_AD ranged from 0.5 to 1.5 during the first harvest of 2016 and from 0.1 to 1.5 in 2017. Wide ranges occurred in subsequent harvests, as well. There was substantial overlap in I_AD values among the three harvests and a lack of a distinct peak in I_AD values in MN and ME. Substantial overlap also occurred in ON in 2016, but a clear peak in I_AD occurred during both harvest dates with 50% and 46% of the fruit having I_AD near 0.7 during the first and second harvests, respectively. A similar peak in I_AD occurred in 2017, but with a clear shift to lower values in later harvests.

Distribution of I_AD values as a percentage of the total harvested ‘Honeycrisp’ apples at three harvest dates in 2016–17. In 2016, the total number of fruit at each harvest in ME was 195, 190, and 214 at harvests 1, 2, and 3, respectively; in MN it was 120, 225, and 225 at harvests 1, 2, and 3 respectively; in ON it was 170 at each harvest. In 2017, the total number of fruit at each harvest in ME was 108, 160, and 126 at harvests 1, 2, and 3, respectively; in MN it was 120 at each harvest; in ON it was 250 at each harvest.

Citation: HortScience 55, 9; 10.21273/HORTSCI15162-20

• Download Figure

View Full Size

Distribution of I_AD values as a percentage of the total harvested ‘Honeycrisp’ apples at three harvest dates in 2016–17. In 2016, the total number of fruit at each harvest in ME was 195, 190, and 214 at harvests 1, 2, and 3, respectively; in MN it was 120, 225, and 225 at harvests 1, 2, and 3 respectively; in ON it was 170 at each harvest. In 2017, the total number of fruit at each harvest in ME was 108, 160, and 126 at harvests 1, 2, and 3, respectively; in MN it was 120 at each harvest; in ON it was 250 at each harvest.

Citation: HortScience 55, 9; 10.21273/HORTSCI15162-20

• Download Figure

Distribution of I_AD values as a percentage of the total harvested ‘Honeycrisp’ apples at three harvest dates in 2016–17. In 2016, the total number of fruit at each harvest in ME was 195, 190, and 214 at harvests 1, 2, and 3, respectively; in MN it was 120, 225, and 225 at harvests 1, 2, and 3 respectively; in ON it was 170 at each harvest. In 2017, the total number of fruit at each harvest in ME was 108, 160, and 126 at harvests 1, 2, and 3, respectively; in MN it was 120 at each harvest; in ON it was 250 at each harvest.

Citation: HortScience 55, 9; 10.21273/HORTSCI15162-20

• Download Figure

Peel I_AD was higher in ‘McIntosh’ than in ‘Honeycrisp’ in all regions. In ‘McIntosh’, region and harvest date affected I_AD with an interaction between the two factors. Peel I_AD decreased with later harvest in each region, but in MN no change occurred between harvests 1 and 2. Regional differences occurred with ME having the highest I_AD, MN having intermediate values, and ON having the lowest.

In ‘McIntosh’, I_AD values within a harvest date were distributed over a narrow range compared with ‘Honeycrisp’ (Fig. 2). Substantial overlap occurred in MN during harvests 2 and 3, and in ON during both harvests. In ME, there was less overlap among harvest dates, and a clear peak in I_AD near 1.9, 1.9, and 1.7 during harvests 1, 2, and 3, respectively.

Distribution of I_AD values as a percentage of all harvested ‘McIntosh’ apples at three harvest dates in 2016. The total number of fruit at each harvest in ME was 195, 135, and 135 at harvests 1, 2, and 3, respectively; in MN it was 120, 270, and 360 at harvests 1, 2, and 3 respectively; in ON it was 200 at each harvest.

Citation: HortScience 55, 9; 10.21273/HORTSCI15162-20

• Download Figure

View Full Size

Distribution of I_AD values as a percentage of all harvested ‘McIntosh’ apples at three harvest dates in 2016. The total number of fruit at each harvest in ME was 195, 135, and 135 at harvests 1, 2, and 3, respectively; in MN it was 120, 270, and 360 at harvests 1, 2, and 3 respectively; in ON it was 200 at each harvest.

Citation: HortScience 55, 9; 10.21273/HORTSCI15162-20

• Download Figure

Distribution of I_AD values as a percentage of all harvested ‘McIntosh’ apples at three harvest dates in 2016. The total number of fruit at each harvest in ME was 195, 135, and 135 at harvests 1, 2, and 3, respectively; in MN it was 120, 270, and 360 at harvests 1, 2, and 3 respectively; in ON it was 200 at each harvest.

Citation: HortScience 55, 9; 10.21273/HORTSCI15162-20

• Download Figure

Fruit maturity.

Chlorophyll a in ‘Honeycrisp’ was related to I_AD in fruit from the first harvest, but not the later harvests, and the range in concentrations was similar in both regions (Fig. 3). Chlorophyll was not measured in ME, nor in 2017 in any location. The range of concentration in ‘McIntosh’ was generally similar to that of ‘Honeycrisp’. Chlorophyll a concentration in ‘McIntosh’ was related to I_AD in MN-grown fruit from the second harvest, but otherwise, it was not related to I_AD. The low r² indicates that I_AD explained less than half the variation in measured chlorophyll a concentration.

The relationship between chlorophyll a concentration in ‘Honeycrisp’ and ‘McIntosh’ fruit peel and the I_AD at harvest. Fruit grown in MN (open symbols) were harvested on three dates, and it was harvested twice in ON (shaded symbols) in 2016. The regression equation for ‘Honeycrisp’ is for both regions and all harvest dates. The regression equation for ‘McIntosh’ is for fruit from the MN second harvest. Each point is a mean of three replications.

Citation: HortScience 55, 9; 10.21273/HORTSCI15162-20

• Download Figure

View Full Size

The relationship between chlorophyll a concentration in ‘Honeycrisp’ and ‘McIntosh’ fruit peel and the I_AD at harvest. Fruit grown in MN (open symbols) were harvested on three dates, and it was harvested twice in ON (shaded symbols) in 2016. The regression equation for ‘Honeycrisp’ is for both regions and all harvest dates. The regression equation for ‘McIntosh’ is for fruit from the MN second harvest. Each point is a mean of three replications.

Citation: HortScience 55, 9; 10.21273/HORTSCI15162-20

• Download Figure

The relationship between chlorophyll a concentration in ‘Honeycrisp’ and ‘McIntosh’ fruit peel and the I_AD at harvest. Fruit grown in MN (open symbols) were harvested on three dates, and it was harvested twice in ON (shaded symbols) in 2016. The regression equation for ‘Honeycrisp’ is for both regions and all harvest dates. The regression equation for ‘McIntosh’ is for fruit from the MN second harvest. Each point is a mean of three replications.

Citation: HortScience 55, 9; 10.21273/HORTSCI15162-20

• Download Figure

Internal ethylene concentration in fruit from ON (not measured in MN or ME) varied with I_AD, but not with harvest date in ‘Honeycrisp’. In both years, IEC was negatively related to I_AD (Fig. 4), so higher IEC occurred with lower I_AD. In 2016, the relationship was the same for the two harvest dates; but in 2017, IEC varied with I_AD during harvests 1 and 2, but not during harvest 3 when the range in I_AD values was narrow. In ‘McIntosh’, IEC was not related to I_AD because most fruit had undetectable levels (not shown).

Internal ethylene concentration (IEC) at harvest of ON-grown ‘Honeycrisp’. IEC is shown as the log of its concentration.

Citation: HortScience 55, 9; 10.21273/HORTSCI15162-20

• Download Figure

View Full Size

Internal ethylene concentration (IEC) at harvest of ON-grown ‘Honeycrisp’. IEC is shown as the log of its concentration.

Citation: HortScience 55, 9; 10.21273/HORTSCI15162-20

• Download Figure

Internal ethylene concentration (IEC) at harvest of ON-grown ‘Honeycrisp’. IEC is shown as the log of its concentration.

Citation: HortScience 55, 9; 10.21273/HORTSCI15162-20

• Download Figure

The starch pattern index did not vary with harvest date when I_AD was included in the statistical analysis, indicating a colinear relationship between the two variables (Table 2). In ‘Honeycrisp’, SPI was linearly and negatively related to I_AD in each region with no interactions, indicating similar relationship to I_AD in the different regions. In MN, the relationship to I_AD was not significant during harvest 1, most likely because of the narrow range in I_AD (Fig. 5) and in ON during harvest 2 when starch breakdown was nearly complete. In ME, SPI was related to I_AD in harvests 1 and 2. In 2017, SPI in ‘Honeycrisp’ was negatively related to I_AD in all regions with no interactions with region or harvest. Within each region, the same relationship between SPI and I_AD occurred for the three harvest dates, but the r² was low in MN. In ME, the relationship was curvilinear because of complete starch breakdown at I_AD values lower than 0.7. In ‘McIntosh’, SPI did not vary among regions and varied little with harvest date and I_AD (Table 2).

Table 2.

Starch pattern index at three harvest dates in ‘Honeycrisp’ and ‘McIntosh’ apples grown in three regions. Starch index was judged on a scale of 1 to 8, with 8 indicating complete breakdown.

View Table

The relationship between starch pattern index at three harvest dates of ‘Honeycrisp’ fruit and fruit peel I_AD at harvest. H1, H2, and H3 indicate harvests 1, 2, and 3, respectively.

Citation: HortScience 55, 9; 10.21273/HORTSCI15162-20

• Download Figure

View Full Size

The relationship between starch pattern index at three harvest dates of ‘Honeycrisp’ fruit and fruit peel I_AD at harvest. H1, H2, and H3 indicate harvests 1, 2, and 3, respectively.

Citation: HortScience 55, 9; 10.21273/HORTSCI15162-20

• Download Figure

The relationship between starch pattern index at three harvest dates of ‘Honeycrisp’ fruit and fruit peel I_AD at harvest. H1, H2, and H3 indicate harvests 1, 2, and 3, respectively.

Citation: HortScience 55, 9; 10.21273/HORTSCI15162-20

• Download Figure

Fruit quality.

Fresh weight, as a measure of fruit size, of ‘Honeycrisp’ in 2016 generally increased with later harvest and was linearly related to I_AD, but with an interaction with region and harvest (Table 3). In MN, fruit weight decreased at higher I_AD values with the second harvest only (not shown). In ME and ON, fruit weight increased at higher I_AD values, but this relationship was only significant for harvest 2 in ON and harvests 1 and 3 in ME. These mixed results indicate that no consistent trend occurred between fruit fresh weight and I_AD. In 2017, fruit weight varied among regions but not other factors. Fruit from MN were smaller than fruit from ON and ME. In ‘McIntosh’, fruit weight was larger in MN than in the other two regions and did not vary with harvest or I_AD.

Table 3.

Mean fresh weight (g) at three harvest dates of ‘Honeycrisp’ and ‘McIntosh’ apples, and red color as a percentage of the peel surface of ‘Honeycrisp’ apples grown in three regions.

View Table

The amount of peel area that had red coloration, measured in ‘Honeycrisp’ fruit in 2017, varied with harvest date and I_AD, but not among regions (Table 3). Peel color generally increased with later harvest and was negatively related to I_AD in all three locations (Fig. 6). Within a region, the linear relationship was similar for all three harvests, although the coefficients of determination varied from 0.4 to 0.8 among regions.

The relationship between red peel color of ‘Honeycrisp’ fruit and I_AD at harvest in fruit from three regions and three harvest dates in 2017.

Citation: HortScience 55, 9; 10.21273/HORTSCI15162-20

• Download Figure

View Full Size

The relationship between red peel color of ‘Honeycrisp’ fruit and I_AD at harvest in fruit from three regions and three harvest dates in 2017.

Citation: HortScience 55, 9; 10.21273/HORTSCI15162-20

• Download Figure

The relationship between red peel color of ‘Honeycrisp’ fruit and I_AD at harvest in fruit from three regions and three harvest dates in 2017.

Citation: HortScience 55, 9; 10.21273/HORTSCI15162-20

• Download Figure

Firmness varied among regions for fruit of ‘Honeycrisp’ (Table 4). In general, firmness was lower in fruit from MN than ON and ME, and it decreased with later harvest. Firmness of ‘Honeycrisp’ fruit at harvest in 2016 was linearly related to I_AD, but not in a consistent or meaningful way (not shown)—sometimes positively, sometimes negatively, and in many instances, with no evident linear relationship. Firmness after storage decreased with later harvest date and was not related to I_AD. ‘Honeycrisp’ fruit firmness in 2017 was not related to harvest date or to I_AD. ‘McIntosh’ fruit firmness in 2016 was not related to harvest date, but firmness after storage varied with I_AD. In fruit from ON, firmness was lower with higher I_AD (y = 12.5 – 2.3x; r² = 0.39), but significant relationships did not occur in MN or ME.

Table 4.

Flesh firmness (N) at harvest and after storage at 0.5 °C of ‘Honeycrisp’ and ‘McIntosh’ apples grown in three regions.

View Table

Soluble solids concentration in ‘Honeycrisp’ fruit at harvest was greater in ME than in ON and MN in 2016 (Table 5). Greater SSC occurred with harvest 2 than with harvest 3. After storage, SSC did not vary with harvest date or region. There was a linear, negative relationship between SSC and I_AD at harvest in MN and after storage in ON (Fig. 7). In ME, SSC was not related to I_AD (not shown). In 2017, SSC was not related to I_AD. In ‘McIntosh’, SSC did not vary among harvests, regions, or I_AD values (Table 5).

Table 5.

Soluble solids concentration (%) at harvest and after 0.5 °C storage of ‘Honeycrisp’ and ‘McIntosh’ apples grown in three regions.

View Table

The relationship between soluble solids concentration (SSC) at harvest and after storage with fruit peel I_AD of ‘Honeycrisp’ fruit from MN and ON and three harvest dates. H1 and H2 indicate harvests 1 and 2, respectively.

Citation: HortScience 55, 9; 10.21273/HORTSCI15162-20

• Download Figure

View Full Size

The relationship between soluble solids concentration (SSC) at harvest and after storage with fruit peel I_AD of ‘Honeycrisp’ fruit from MN and ON and three harvest dates. H1 and H2 indicate harvests 1 and 2, respectively.

Citation: HortScience 55, 9; 10.21273/HORTSCI15162-20

• Download Figure

The relationship between soluble solids concentration (SSC) at harvest and after storage with fruit peel I_AD of ‘Honeycrisp’ fruit from MN and ON and three harvest dates. H1 and H2 indicate harvests 1 and 2, respectively.

Citation: HortScience 55, 9; 10.21273/HORTSCI15162-20

• Download Figure

Storage disorders.

Incidence of bitter pit in ‘Honeycrisp’ fruit was not related to I_AD at harvest in either year and did not vary with harvest date in MN or ME. Bitter pit incidence in 2016 in ME was 11%, 8%, and 5%; and it was 0%, 2%, and 0% in MN for the first, second, and third harvests, respectively. In ON, bitter pit did not occur in 2016, but it was severe in 2017 when incidence was 7%, 13%, and 31% for each harvest, respectively. In 2017, bitter pit was less severe in ME than in the previous year, and it was 2%, 0%, and 0.4% for each respective harvest. In MN, bitter pit occurrence in 2017 was similar to that observed in 2016: 0%, 0%, and 2% for each respective harvest. The occurrence of diffuse flesh browning varied with year and harvest date, but it was not related to I_AD (data not shown). Diffuse flesh browning was more prevalent in 2017 than in 2016 in both ON and ME, where it was greater in fruit from the third harvest than in the first or second.

Soft scald occurred in ‘Honeycrisp’ fruit in all three regions in both years (Table 6). Fruit were not conditioned before cold storage to induce full expression of the disorder. Incidence in 2016 was lower in MN and ON than in ME, where it was severe. There was no relationship with I_AD in 2016 in MN and ON (Fig. 8). In ME, SSI was related to I_AD, but the relationship varied considerably between the three harvests. Soft scald was negatively related to I_AD in fruit from harvest 1; positively related to I_AD in fruit from harvest 2; and not related to I_AD with harvest 3, when it was severe for all I_AD categories. In 2017, SSI was not related to I_AD in MN. Fruit from harvest 1 had greater SSI than later harvests in MN, as well as in ON. In ON, SSI was positively related to I_AD with all three harvests. In ME, the relationship to I_AD occurred in a similar manner as in 2016, when SSI was negatively related to I_AD for the first harvest and positively related with I_AD values for harvests 2 and 3.

Table 6.

Soft scald and soggy breakdown incidence after 0.5 °C storage of ‘Honeycrisp’ apples grown in three regions for two years.

View Table

The relationship between soft scald incidence in ‘Honeycrisp’ and fruit peel I_AD at harvest in three locations and 2 years. H1, H2, and H3 indicate harvests 1, 2, and 3, respectively.

Citation: HortScience 55, 9; 10.21273/HORTSCI15162-20

• Download Figure

View Full Size

The relationship between soft scald incidence in ‘Honeycrisp’ and fruit peel I_AD at harvest in three locations and 2 years. H1, H2, and H3 indicate harvests 1, 2, and 3, respectively.

Citation: HortScience 55, 9; 10.21273/HORTSCI15162-20

• Download Figure

The relationship between soft scald incidence in ‘Honeycrisp’ and fruit peel I_AD at harvest in three locations and 2 years. H1, H2, and H3 indicate harvests 1, 2, and 3, respectively.

Citation: HortScience 55, 9; 10.21273/HORTSCI15162-20

• Download Figure

Soggy breakdown incidence in ‘Honeycrisp’ varied with region, harvest, and I_AD with an interaction among all three factors in both years (Table 6). In ON, soggy breakdown did not occur. In MN, soggy breakdown was greater with harvest 3 than the other harvests in 2016; did not vary with harvest date in 2017; and did not vary with I_AD (Fig. 9). In ME, soggy breakdown was related to I_AD in a similar manner as with SSI in 2016. In 2017, soggy breakdown was related to I_AD with harvest 3 and not with the previous harvests.

The relationship between soggy breakdown incidence in ‘Honeycrisp’ fruit and fruit peel I_AD at harvest in MN and ME and 2 years. Soggy breakdown did not occur in ON apples. H1, H2, and H3 indicate harvests 1, 2, and 3, respectively.

Citation: HortScience 55, 9; 10.21273/HORTSCI15162-20

• Download Figure

View Full Size

The relationship between soggy breakdown incidence in ‘Honeycrisp’ fruit and fruit peel I_AD at harvest in MN and ME and 2 years. Soggy breakdown did not occur in ON apples. H1, H2, and H3 indicate harvests 1, 2, and 3, respectively.

Citation: HortScience 55, 9; 10.21273/HORTSCI15162-20

• Download Figure

The relationship between soggy breakdown incidence in ‘Honeycrisp’ fruit and fruit peel I_AD at harvest in MN and ME and 2 years. Soggy breakdown did not occur in ON apples. H1, H2, and H3 indicate harvests 1, 2, and 3, respectively.

Citation: HortScience 55, 9; 10.21273/HORTSCI15162-20

• Download Figure

Disorders did not occur in ‘McIntosh’ grown in ME or ON. ‘McIntosh’ fruit grown in MN developed core browning and soft scald. There was a weak, positive relationship between core browning incidence and I_AD with harvests 2 and 3 (r² = 0.26). At I_AD values below 1.5, no core browning occurred; but above this, core browning increased to 20% at an I_AD of 1.7 (not shown). No core browning occurred with harvest 1. Soft scald occurred in 5% of the ‘McIntosh’ apples grown in MN, but incidence was not related to I_AD (not shown).

Discussion

The I_AD should change in a measurable way and in parallel with later harvest date or advancing maturity. This has been shown to occur in ‘Ambrosia’ (Toivonen, 2015), ‘Aurora Golden Gala’ (Toivonen and Hampson, 2014), ‘Empire’ (Doerflinger et al., 2016), ‘Honeycrisp’ (DeLong et al., 2014) and ‘Starking’ (Nyasordzi et al., 2013) fruit. In this study, we found a reduction in I_AD values with later harvest dates in ‘Honeycrisp’ and ‘McIntosh’ fruit, but sometimes differences were slight, particularly in ‘McIntosh’ fruit. I_AD values of ‘Honeycrisp’ fruit were similar to previous reports (DeLong et al., 2014; Serra et al., 2016), but none are available for ‘McIntosh’ fruit.

In this study, fruit were randomly harvested to estimate the range in I_AD within a harvest date. A notable difference between the two cultivars was the wider range in I_AD values in ‘Honeycrisp’ fruit, both within harvest date and during the harvest period, suggesting that ripening is more uneven in ‘Honeycrisp’ than in ‘McIntosh’. An additional difference was the higher I_AD values in ‘McIntosh’ compared with ‘Honeycrisp’ fruit, despite somewhat similar chlorophyll a concentration. This occurs in ‘Granny Smith’ and ‘Starking’, as well, where both had similar chlorophyll concentration but widely varying I_AD (Nyasordzi et al., 2013). The high I_AD values and distribution of I_AD values around a distinct peak in ‘McIntosh’ fruit was similar to the distribution for ‘Granny Smith’ fruit, whereas the wider range in values without the distinct peak in ‘Honeycrisp’ fruit was similar to that in ‘Cripps Pink’ fruit (Farneti et al., 2015) and previously reported values for ‘Honeycrisp’ (Serra et al., 2016).

The stage of maturity that determines when to harvest should be indicated by the same I_AD from year to year, and it should be consistent with postharvest performance. In ‘Ambrosia’ fruit, I_AD is more consistent from year to year than SPI in determining when to harvest for firmness retention in storage (Toivonen, 2015). We found year-to-year consistency in I_AD in ‘Honeycrisp’ between ON and in ME during harvest 2 in year 1, but not at other times. In MN, I_AD values were higher in year 1 compared with year 2, but starch breakdown was also less advanced in year 1 compared with year 2. Some of these year-to-year differences may be due to inconsistency in judging when to start harvest.

Orchard variability in I_AD occurs in ‘Honeycrisp’ fruit when measured within a region (DeLong et al., 2014; Serra et al., 2016), but not in ‘Starking’ or ‘Pink Lady’ (Nyasordzi et al., 2013). We measured only one orchard in each region. Similarity in the range of I_AD values during each harvest occurred among regions in most cases. However, regional inconsistencies occurred that could not be explained, such as I_AD values that were twice as high in ME as in ON in 2017 despite greater starch breakdown in ME. Although the starch index can be influenced by cropload, temperature, and plant stress (Smith et al., 1979), it is often used as one indicator of fruit maturity. Small regional differences also occurred with ‘McIntosh’ that were not consistent with SPI.

A strong relationship occurs between I_AD and chlorophyll concentration when measured over a 9-week (DeLong et al., 2014) or 5-week (Shao et al., 2014) range in harvest dates. The correlation occurs in many cultivars (Shao et al., 2014), but it did not occur in our study with only a 2-week range in harvest dates, which may have been too narrow to ensure a sufficiently wide range in I_AD values. In addition, the two cultivars had similar chlorophyll concentrations, but different ranges in I_AD. This occurs in ‘Granny Smith’ and ‘Starking’, as well, which had similar chlorophyll concentration, but I_AD values that differed by 0.9 units (Nyasordzi et al., 2013).

In ‘Honeycrisp’ fruit, I_AD was related to the maturity indicators IEC and SPI in both years, and to SPI to different degrees in all three regions. However, the relationship to SPI was not the same for each harvest date in 2016 because starch breakdown occurred faster than reduction in I_AD and was nearly complete by the third harvest. I_AD correlation with peel blush occurred in ‘Honeycrisp’ fruit in all three regions and with the same relationship for all three harvests, indicating the two traits are closely linked under the growing conditions in this study. However, the relationship of peel blush and I_AD in MN was weaker and with a smaller slope in MN than in ON and ME.

In this study, the region where fruit were grown and stored was more important in determining susceptibility to SSI than harvest maturity. However, only one orchard was sampled within each region, so conclusions regarding regional variation may be because of differences between the three orchards. Soft scald incidence was sometimes more severe in fruit harvested earlier rather than later, as in the case of MN and ON in 2017, indicating that SSI is not consistently more severe in fruit with advanced maturity. A decrease in SSI with later harvest date was previously reported for Michigan-grown ‘Honeycrisp’ (Watkins et al., 2005). In ME, SSI had a negative relationship to I_AD during the first harvest, indicating an increase in susceptibility with advanced maturity. However, the relationship of SSI and I_AD switched from negative to positive during the harvest window in ME as mean I_AD decreased below 0.8, as in ON in 2017 where it was positive for all three harvests. This positive relationship to I_AD also occurred in the same orchard in 2015 with the latter harvest when SSI was severe (Moran, 2018). This indicates a consistent year-to-year trend in ME, where SSI is typically severe. In these cases, the relationship to I_AD was positive, indicating reduced susceptibility in fruit with more advanced maturity when examining fruit from within a harvest date. Al Shoffe et al. (2020) found no linear correlation between I_AD and SSI in fruit grown in Pennsylvania and New York. Harvest date has a large, but inconsistent effect on ‘Honeycrisp’ fruit disorder occurrence, such that the same I_AD value may result in a low incidence of SSI for one harvest, but a high incidence for a different harvest. ‘Honeycrisp’ fruit ripening and disorder development seems to be sensitive to environmental conditions (Lachapelle et al., 2013; Moran et al., 2009), so there can be much variation in the timing of various fruit ripening characteristics (change in background color, ethylene production, rate of starch loss, etc.) within and among orchards.

Conclusions

I_AD was useful as an adjunct to other measures of maturity in ‘Honeycrisp’ fruit, with some consistency in the range in values among two of three regions in each year. The greatest degree of regional disparity in I_AD values occurred with the first harvest but was reflective of disparity in SPI among the regions during harvest 1. I_AD measured changes in ‘McIntosh’ fruit maturity in one of the three regions in this study, but the decrease in I_AD with later harvest was too small in the other two regions. This may reflect the use of different sports of ‘McIntosh’ grown at the different regions. Incidence of bitter pit was not predicted by I_AD at harvest, whereas SSI was predicted by I_AD, but only in two regions and in contrasting ways from harvest to harvest. Due to regional and yearly inconsistencies, growers should use I_AD data as one of several methods, such as SPI and change in ground color, to judge ‘Honeycrisp’ apple maturity. I_AD was not useful for measuring maturity in ‘McIntosh’ apples but was weakly related to core browning incidence.