Pre-Harvest Pneumatic Defoliation and Pruning Improves Peel Coloration of ‘Nicoter’ and ‘Rosy Glow’ Apples Depending on Exposition, Fruit Side and Weather Conditions

Abstract

Peel coloration in bicolored apples is decisive for their commercial value and varieties like ‘Nicoter’ and ‘Rosy Glow’ need to achieve 33 and 40% red peel color to be marketed under the trade names ‘Kanzi^®’ and ‘Pink Lady^®’, respectively. Pneumatic defoliation of apple trees is a novel technique to lower the amount of leaf area and, along with pre-harvest pruning, it can improve peel coloration by increasing sunlight incidence on fruit. Our experiments showed that in fact the color index and the amount of anthocyanins, the latter being responsible for the red fruit peel, increased up to 100% after pneumatic defoliation (PD) and pre-harvest pruning (P), or their combination (P+PD), over three years. Anthocyanin contents in ‘Rosy Glow’ apple peels rose irrespective of the orientation and side of the fruit, while in ‘Nicoter’ they increased only on the sun-exposed fruit side. The extent depended on the harvest year and its meteorological conditions and our results showed that the effect is greater in years with high pre-harvest temperatures. In most combinations of year and variety, the three treatments P, PD and P+PD were equally effective.

1. Introduction

Bicolored apples are appreciated on the market for their defined ratio of red to yellow–green skin color. Among bicolored cultivars, ‘Nicoter’ is a cross between Gala and Braeburn, while ‘Rosy Glow’, a mutation of ‘Cripps Pink’, is a hybrid of Lady Williams and Golden Delicious. To be sold under the trade names ‘Kanzi^®’ and ‘Pink Lady^®’, the red color on the fruit of cv. ‘Nicoter’ and ‘Rosy Glow’ must cover at least 33 and 40%, respectively, of the peel area. The red coloration of apple skin is due to anthocyanins, pigments produced under specific environmental conditions. The main anthocyanins in apple peels are cyanidin-3-O-galactoside, cyanidin-3-O-glucoside and cyanidin-3-O-arabinoside [1]. Their synthesis is influenced not only by intense solar radiation and low temperatures [2] but also by phytohormones and gene transcription factors, which vary in expression across different apple varieties. Anthocyanin synthesis includes a series of biochemical transformations involving various regulatory genes, the most important belonging to the myeloblastosis (MYB) family [3]. MYB proteins are responsible for a variety of plant processes, such as their stress response, development, cell shape and metabolism [4]. While genetic modifications of new apple varieties interfere directly at the basis of anthocyanin production in the plant, orchard management practices aim at optimizing environmental conditions for anthocyanin production in the fruit skin [5]. The temperature regime surrounding the fruit mainly depends on the location and the topography of the apple orchard and on annual weather conditions. Light incidence on the fruit can be manipulated by cultivation practices, for example, through the use of different training systems, by increasing the albedo with reflective material on the orchard floor or by leaf removal through pre-harvest pruning (P) or pneumatic defoliation (PD) [6,7]. Andergassen et al. have extensively studied the effects of leaf removal by P and PD on the amount of solar radiation penetrating the canopy [8]. In that work, the consequences of P and PD on several fruit quality parameters were evaluated [8]. However, anthocyanin changes on the fruit skin were discussed only for a single harvest year and measured on the red, sun-exposed fruit side, without accounting for possible effects of the defoliation treatments on the sun-averted fruit side or for differences between fruits grown on sides of the trees with different expositions (east and west). Yet cardinal directions can be decisive, especially since the incidence of morning and evening sun on fruit can differ considerably in duration and strength. Generally, studies considering the impact of the treatments during several years with different climate conditions have to be performed to understand the influence of defoliation treatments under varying environmental conditions. This might give useful information on the conditions and extent of the treatment in routine application, e.g., which of the treatments, P, PD or P+PD, is the most efficient and under which meteorological conditions it should be applied. This study, built on the work of Andergassen et al. [8], aims to close this gap by examining the effects of defoliation treatments (P, PD and their combination P+PD) on the color index and anthocyanin content of ‘Nicoter’ and ‘Rosy Glow’ apple peels, depending on their position on the three, the exposition and the fruit side. We hypothesized that defoliation effects vary with the fruit position within the canopy. We further assessed the effect of the treatments in relation to the global radiation and the temperatures during the pre-harvest period in the different sampling years.

2. Materials and Methods

2.1. Plant Material

Trials were carried out on Malus domestica Borkh. ‘Nicoter’ planted in 2007 and ‘Rosy Glow’ planted in 2012 and managed as described in Andergassen et al. [8]. Trees were grown as slender spindles on M9 T337 rootstocks, planted in a north–south orientation, in orchards located at the Laimburg Research Centre, South Tyrol, Italy (‘Rosy Glow’: 46°23′09.5″ N, 11°17′33.2″ E; ‘Nicoter’: 46°20′37.2″ N, 11°16′43.0″ E) at altitudes of 220 and 223 m above sea level, respectively, for ‘Nicoter’ and ‘Rosy Glow’. Tree spacing was 3.0 m × 0.8 m (‘Nicoter’) and 3.3 m × 1.0 m (‘Rosy Glow’), with a tree height of ca. 3.5 m. Both cultivars were managed as club varieties according to GLOBALG.A.P. practices [9]. A randomized complete block design with four blocks and four treatments was employed for each variety. The treatments were carried out in 2021, 2022, and 2023 and consisted of an untreated control (UTC), in which no leaf removal was employed, application of pre-harvest pruning (P) or pneumatic defoliation (PD), either alone or in combination (

Table 1. Pre-harvest pruning (P), pneumatic defoliation (PD) and fruit harvest dates for ‘Nicoter’ and ‘Rosy Glow’ varieties used in the experiments. Only the harvest date of the first of three picks is indicated.

Event	‘Nicoter’				‘Rosy Glow’
	2021	2022	2023		2021	2022	2023
Pre-harvest pruning (P)	24.08.	30.08.	29.08.		07.10.	03.10.	13.10.
Pneumatic defoliation (PD)	03.09.	05.09.	04.09.		07.10.	11.10.	19.10.
Fruit harvest date	17.09.	14.09.	14.09.		27.10.	26.10.	26.10.

).

In summary, the following treatments were carried out on trees of ‘Nicoter’ and ‘Rosy Glow’:

-

UTC (untreated control, no pruning or defoliation)

-

P (pre-harvest pruning)

-

PD (pneumatic defoliation)

-

P+PD (pre-harvest pruning and pneumatic defoliation).

P consists in the manual removal of water sprouts and unproductive long shoots, carried out once per season as reported in Table 1, while PD was performed with a pneumatic leaf blower (Vortex, Olmi, Asti, Italy). Two passages were carried out at a speed of 1 km h⁻¹ and a pressure of 0.8 bar, one on the lower and one on the higher part of the canopy (maximum height ca. 3.0 m). For ‘Nicoter’, P removed on average 18%, PD 21% and P+PD 49% of the leaf area, while for ‘Rosy Glow’, P reduced the leaf area by 30%, PD by 36% and P+PD by 60% on average [8] (Andergassen et al. 2023). Treatments were performed each year on the same trees. The percentage of red color on the apple peels was analyzed by a commercial grading machine (Calistar, AWETA, Pijnacker, The Netherlands). All apples from the first pick were considered for the calculation of apples with more than 33% (‘Nicoter’) and 40% (‘Rosy Glow’) red peel color. An experimental unit consisted of a 16 m long block along one row, amounting to 20 and 16 adjacent trees for ‘Nicoter’ and ‘Rosy Glow’, respectively, due to the different variety-specific plant distances of 0.8 and 1.0 m, respectively. Three trees per experimental unit and cultivar were selected randomly for color index and anthocyanin assessment. For ‘Nicoter’ and ‘Rosy Glow’, a common practice consists of harvesting fruit in various (between 1 and 4) picks throughout a period of up to one month, through the timely selection of mature fruit with each pick without retarding the harvest of already mature apples and avoiding picking premature ones. For our experiments, four apples per tree were collected at a canopy height of 1.0–1.5 m, considering only the first of three picks (Table 1). The considered harvest dates in each year 2021, 2022 and 2023 are given in Table 1 and were determined by assessment of the starch index [10]. Two apples each were picked from the west and from the east side, one of which was from the inside of the canopy, where the tree crown shields the fruit, and the other from the outside of the canopy, i.e., from the final part of the shoots (Figure 1).

Accordingly, 192 fruit samples per cultivar (4 defoliation treatments × 4 trees × 3 apples × 2 expositions × 2 canopy positions) per year were collected for analytical measurements of the color index and the peel anthocyanin contents. Apples were analyzed on the sun-exposed and on the sun-averted fruit side Each apple and its fruit side was measured in duplicate for its color index (total of 768 measurements per year and variety), and the three apples per tree position and the duplicate measurements were then averaged for data evaluation. The three apples of each position (Figure 1, grey circles) were pooled for the measurement of the anthocyanin content (total of 128 measurements per year and variety).

2.2. Weather Data

Data were assessed throughout the three sampling years, 2021, 2022 and 2023, at a weather station placed at the site of the ‘Rosy Glow’ orchard (46°22′56.9″ N, 11°17′19.5″ E). The ‘Nicoter’ orchard was situated at a 4.7 km straight-line distance from the weather station, experiencing similar weather conditions. Global solar radiation and air temperature minima and maxima were depicted only for the last 14 days before harvest of both varieties in each sampling year (see Table 1). For ‘Nicoter’, the last 14 days before harvest were from 31 August to 14 September in 2022 and 2023 and from 3 to 17 September in 2021. For ‘Rosy Glow’, the last 14 days before harvest included the period from 12 to 26 October in 2022 and 2023 and from 13 to 7 October in 2021. The incoming global solar radiation was calculated as a daily sum of the values measured every 10 min from 0:00 to 24:00 h with a CMP6 pyranometer with a CVF3 ventilation unit (Kipp&Zonen, Delft, The Netherlands). The air temperature was measured under air circulation in the shade every 10 min at a height of 50 cm from the ground with a Hygroclip HC2-S3 sensor with RS12T radiation protection (Rotronic, Passirana di Rho, Italy), considering only the daily minima and maxima for evaluation. Due to its importance for color formation the daily temperature range (max–min) was calculated and is reported.

2.3. Assessment of the Percentage of Red Color and Calculation of the Color Index (CI)

The color index (CI) can be defined as the ratio between the intensities of red, yellow, green and blue colors. It is a direct measure for the coloration of apple peels, which is in turn caused by their anthocyanin content [11]. The CI was measured with a spectrophotometer (CR-400, Konica Minolta, Nieuwegein, The Netherlands) against a white standard on the sun-exposed and sun-averted fruit side, each in duplicate, of each apple. Values for the sun-exposed fruit side were obtained by measuring the most intensely red-colored part of the peel, while those on the sun-averted fruit side were taken on its counterpart (yellow–green-colored side of the peel). The values were calculated as an average of the two measurements and three replicates according to Formula (1) [12]:

CI = 1000 × Δa/(Δb × ΔL)

(1)

Δa is the red/green difference, Δb the yellow/blue difference and ΔL the lightness/darkness ratio.

2.4. Measurement of the Anthocyanin Content

Anthocyanin content in the apple peel was analyzed following the method described by Ceci et al. [13]. A 3 cm × 3 cm peel slice (ca. 2.5 g) was removed from the same apples used for the CI determination with a commercial potato peeler, removing the least possible amount of pulp. All slices were peeled by the same operator to avoid systematic errors. The slices were lyophilized (FreeZone Freeze Dry System, Labconco, MO, USA) and milled to a fine powder using a ball mill (MM400, Retsch, Haan, Germany). An aliquot of 25 mg powder was weighed, and 1.8 mL solution A (0.5 g concentrated phosphoric acid in 40 mL methanol and 10 mL MilliQ water) and 30 µL solution B (0.21 g sodium fluoride in 50 mL MilliQ water) were added. The mixture was shaken at 5 °C/750 rpm for 15 min (Thermomixer, Eppendorf, Hamburg, Germany) and then centrifuged at 5 °C/12,000 rpm for 5 min (Centrifuge 5810 R, Eppendorf, Hamburg, Germany). One sample aliquot of 200 µL was mixed with 800 µL solution C (54.43 g sodium acetate in 1 L MilliQ water, pH 4.5 adjusted with concentrated hydrochloric acid) and another 200 µL aliquot with 800 µL solution D (1.86 g potassium chloride in 1 L MilliQ water, pH 1.0 adjusted with concentrated hydrochloric acid). The anthocyanin content was determined by reading the absorbance at 520 and 700 nm by UVVIS spectrometry (Cary 60, Agilent, Santa Clara, CA, USA). Total anthocyanins were calculated as cyanidin-3-O-glucoside equivalents, since the latter is one of the most abundant anthocyanins in apple peels [1]. The following formula was used (2):

$$\mathrm{total} \mathrm{anthocyanins}={\displaystyle \frac{{(\mathrm{A}}_{520 \mathrm{nm}}-{\mathrm{A}}_{700 \mathrm{nm}}{)}_{\mathrm{pH}1.0}\times 449.2{ \mathrm{g} \mathrm{mol}}^{-1}\times 5\times \mathrm{100,000}\times 0.93}{{\mathrm{26,900} \mathrm{g} \mathrm{mol}}^{-1} {\mathrm{cm}}^{-1}\times 1 \mathrm{cm}\times \mathrm{sample} \mathrm{weight} \left[\mathrm{mg}\right]}}$$

(2)

Anthocyanin contents were expressed in mg 100 g⁻¹ fresh weight.

2.5. Statistical Analysis

The relationship between anthocyanin content and CI was investigated by linear regression, accounting for the anthocyanin content as the independent variable and the CI as the dependent variable. For each apple cultivar and year, the effect of the defoliation treatments on the percentage of red peel color was analyzed by analysis of variance. Normal distribution of the data was visually checked by diagnostic plots [14]. If data met the assumptions of normality and variance homogeneity, analysis of variance (type III sum of squares), accounting for the defoliation treatment and the block, was conducted (‘Nicoter’ cultivar: 2021, 2022, 2023; ‘Rosy Glow’ cultivar: 2023). If the assumptions (normal distribution of residuals and variance homogeneity) were violated, a non-parametric test (Kruskal–Wallis), accounting for the defoliation treatment only, was used (‘Rosy Glow’ cultivar: 2021, 2022). The effect of the block, the defoliation treatment (DT), the exposition of the apple on the tree (EX), the position of the apple in the canopy (PO) and the fruit side (FS), as well as the interactions between the four latter factors, all treated as fixed terms, on the anthocyanin content were analyzed by linear mixed models. The plot was included as a random term to account for correlations between measurements corresponding to the combinations of different levels of DT, EX and PO within the same experimental unit. The FS was treated as a factor involving repeated measurements on the single fruit, from which values of the sun-exposed and the sun-averted fruit side were measured, as a subject. Type III sum of squares with Satterthwaite correction of the denominator degrees of freedom and Restricted Maximum Likelihood estimation were used. Fulfillment of the assumptions (normal distribution of residuals and homoscedasticity) was visually assessed by diagnostic plots [14], and natural logarithm–transformation was performed for the anthocyanin values to meet these assumptions. Planned multiple comparisons were performed at the highest significant order of interaction involving DT by Least Significant Difference (LSD). A p-value < 0.05 was considered significant. All analyses were carried out by IBM SPSS Statistics version 29.0.1.0 171. The linear regression was conducted in SPSS and visualized using Microsoft Excel version 2407.

3. Results

3.1. Meteorological Conditions

The last 14 days before harvest are a crucial period for anthocyanin formation in apple peels [15]. Each year, the global radiation in this period showed some variability between one day and the next and tended to be lower before ‘Rosy Glow’ than before ‘Nicoter’ harvest (Figure 2). Although trees were grown in the same geographical area, ‘Rosy Glow’ was harvested, on average, 41 days later than ‘Nicoter’, towards the end of summer. That is, ‘Nicoter’ was harvested in late summer, between 14 and 17 September depending on the harvest year, while ‘Rosy Glow’ was harvested in autumn, on 26 and 27 November depending on the harvest year (see Table 1). The inherent differences in fruit maturity and harvest dates between the two varieties account for differences in radiation and temperature characteristics for the respective season of the year. During the last five days before the ‘Rosy Glow’ harvest, the solar radiation was generally highest in 2021, while in 2023 relatively low solar radiation was recorded just before harvest. Average solar radiation in the two weeks before harvest amounted to 24.8, 24.7 and 28.2 MJ m⁻² for ‘Nicoter’ and 15.3, 12.6 and 10.8 MJ m⁻² for ‘Rosy Glow’ in 2021, 2022 and 2023, respectively (see Supporting Information Figures S1 and S2).

Similarly, the minimum temperatures in the last 14 days before harvest were different in the three sampling years for both varieties, ‘Nicoter’ and ‘Rosy Glow’ (Figure 3A). The lowest values were obtained for both in 2021 (on average, 12.2 and 1.4 °C, respectively), while 2022 and 2023 showed similar higher values (14.1 and 14.4 °C, respectively, for ‘Nicoter’ and 8.7 and 9.0 °C, respectively for ‘Rosy Glow’). Daily temperature minima were reached between 6 a.m. (‘Nicoter’) and 7 a.m. (‘Rosy Glow’) on average. Maximum temperatures (Figure 3B) before ‘Rosy Glow’ harvest were highest in 2022 (22.5 °C on average) in comparison to 2021 (19.4 °C) and 2023 (20.3 °C). Before harvest of ‘Nicoter’, maxima were considerably higher than in ‘Rosy Glow’ in all three years (30.1, 29.4 and 31.5 °C in 2021, 2022 and 2023, respectively), as expected because of the harvest occurring at an earlier stage of the summer season. Daily temperature maxima were reached on average at 3 pm. Accordingly, daily temperature excursion (see Supplementary Information Figures S3–S6), decisive for anthocyanin formation, in the last 14 days before harvest was greatest in 2021 for ‘Rosy Glow’.

3.2. Correlation Between the CI and the Anthocyanin Contents

The CI in apple peels of ‘Nicoter’ exhibited a close relationship with their anthocyanin contents in all three sampling years and for all treatments (p < 0.001; coefficients of determination R² over all data points 0.97, 0.95 and 0.96, respectively, for 2021, 2022 and 2023). A clear differentiation between values measured on the sun-averted fruit side (Figure 4A, black dotted line) and on the sun-exposed fruit side was visible, with the latter showing a distinction between apples grown inside (Figure 4A, dark grey dashed line) and outside (Figure 4A, light grey solid line) the tree canopy, irrespective of the applied treatment. CIs measured on the sun-averted fruit side exhibited low or slightly negative values (between 1.8 and −4.2) due to the small red color component with respect to the yellow–green proportion. On the sun-exposed fruit side, apples grown outside the canopy had the highest CIs with values up to 46.1 compared to a maximum of 29.6 for apples grown inside the canopy. In 2021, the distinction on the sun-exposed fruit side between apples grown inside and outside the canopy was less obvious due to slight overlaps of the elevated P+PD values inside the canopy with those from outside the canopy (Figure 4A, left). Especially in 2022, CI and anthocyanin values on the sun-exposed fruit side were considerably higher than those on the sun-averted fruit side, while the distinction on the sun-exposed fruit side of apples grown inside and outside the canopy was less evident, but still clearly visible (Figure 4A, center). In 2023, all three groups (sun-averted fruit side, sun-exposed fruit side from apples inside and outside the canopy) were clearly separated and no overlaps occurred, independent of the treatment (Figure 4A, right). The explicit distinction of CI and anthocyanin values in the stated groups without apparent separation of the values from different defoliation treatments evidences the necessity of statistically analyzing the efficiency of the defoliation treatments, taking into consideration not only the different fruit sides and canopy positions but also the fruit exposition (grown on the east or west side of the tree) and the interaction between all the mentioned parameters.

For ‘Rosy Glow’, the CI of the apple peels also correlated well with their anthocyanin content for all three sampling years independent of the treatment (p < 0.001; R² over all data points 0.96, 0.96, and 0.99, respectively, in 2021, 2022 and 2023). It must be noted that the CI and anthocyanin values were generally higher than in ‘Nicoter’, with CI maxima of 55.2 and anthocyanin contents of 36.53 mg 100 g⁻¹ fresh weight (Figure 4B), indicating a more pronounced red coloration of the cv. ‘Rosy Glow’, in line with different standards for marketing the varieties. Positive CI values were observed not only on the sun-exposed but also on the sun-averted fruit side, with exceptions in 2023, where values of fruit grown inside the canopy were partly negative and as low as −2.8. When considering only the apples grown inside (dark grey dashed line) and outside (light grey solid line) the canopy throughout the entire dataset, no clear separation between the values measured on the sun-averted (black dotted lines) and those of the sun-exposed fruit side was apparent in 2021 (Figure 4B, left). However, when UTC, P, PD, and P+PD were examined individually, CIs and anthocyanins on the sun-averted and sun-exposed fruit side from apples outside and inside the canopy evidently differed between these groups. In contrast to ‘Nicoter’, a group distinction was even possible on the sun-averted fruit side between apples from inside and outside the canopy by regarding the single treatments. In 2022 the difference between values from inside and outside the canopy, both on the sun-averted and the sun-exposed fruit sides, was even more evident and visible without considering the treatments singularly (with exceptions for the low values of the UTC sun-exposed fruit side outside the canopy and few overlaps between sun-averted and sun-exposed fruit side inside the canopy, Figure 4B, center). In 2023, a very clear distinction was noted throughout all treatments between the sun-averted and sun-exposed fruit side, inside and outside the canopy (Figure 4B, right). This group distinction of the different treatments, fruit sides and canopy positions made a statistical examination of possible interactions between these factors advisable.

3.3. Impact of the Defoliation Treatments on the Anthocyanin Contents

Anthocyanin contents in peels of ‘Nicoter’ fruits sampled in 2021 increased by the defoliation treatments. This effect depended, according to a three-way interaction, on the fruit side and exposition (p < 0.001, F = 19.5, see Supporting Information Table S1). PD and P+PD enhanced the anthocyanin contents significantly on the sun-exposed fruit side for apples grown inside the canopy and on the sun-averted fruit side for those grown outside the canopy (Figure 5 left). An anthocyanin increase is thus visible only for fruit obtaining medium sunlight intensity, and not for those with high (sun-exposed fruit side of apples from outside the canopy) or low (sun-averted fruit side of apples inside the canopy) solar radiation. The increase in anthocyanins on the sun-exposed fruit side stands in line with a significant increase in the percentage of fruit with >33% red peel color, an important prerequisite for marketing the variety under its trade name ‘Kanzi^®’. While UTC samples consisted of 97% fruit with more than 33% red color, all defoliation treatments slightly enhanced the percentage (Figure 6 left). In 2022, treatments were unable to enhance the peel anthocyanins in ‘Nicoter’ apples (Figure 5 center), irrespective of the fruit side and exposition (see Supporting Information Table S1), which could be attributed to the fact that the anthocyanin content in the UTC was higher than in the other sampling years (9.62 mg 100 g⁻¹ fresh weight compared to 6.00 mg 100 g⁻¹ fresh weight in 2021 and 7.45 mg 100 g⁻¹ fresh weight in 2023, see Supporting Information Figure S7). Despite the lack of influence of the treatments on the anthocyanin content, the percentage of fruit with >33% red color increased significantly from 83% to 91 and 94%, respectively, upon PD and P+PD, both showing a similar effect (no significant differences between PD and P+PD, although P+PD was significantly higher than P, while PD was not, Figure 6 center). The increasing percentage might be explained by an enhanced spread of anthocyanins over the entire peel, while the sampled peel slices (most red-colored part of the peel and its counterpart) lacked intensification. In 2023, we found a two-way interaction between the defoliation treatment and the fruit side (F = 3.218, p = 0.031, see Supporting Information Table S1), impacting the anthocyanin content. While all treatments (P, PD and P+PD) enhanced the anthocyanin contents with respect to the UTC on the sun-exposed fruit side, no changes were observed on the sun-averted fruit side (Figure 5). PD increased the anthocyanins most, i.e., by 45% compared to the UTC (P by 32% and P+PD by 23%). Although the anthocyanin contents increased by all treatments, the percentage of apples with >33% red color increased significantly from 77% in the UTC to 91% only following P+PD, while PD exhibited intermediate values (82%) not differing significantly from UTC and P on one side and from P+PD on the other side (Figure 6 right).

Anthocyanin contents in ‘Rosy Glow’ increased in 2021 with all defoliation treatments (Figure 6), irrespective of the fruit side and its exposition and canopy position (see Supporting Information Table S2). In 2022, a two-way interaction of the treatment with the exposition of the apple on the tree was detected (F = 3.523, p = 0.019, see Supporting Information Table S2). All treatments enhanced the anthocyanin contents significantly, but only in fruit peels of apples grown on the east side of the tree. In accordance with what was found for ‘Nicoter’ fruits, the anthocyanin contents for the UTC of ‘Rosy Glow’ were on average higher in 2022 (11.57 mg 100 g⁻¹ fresh weight) than in 2021 and 2023 (10.50 and 8.83 mg 100 g⁻¹ fresh weight, respectively, see Supporting Information Figure S8), possibly accounting for the anthocyanin increase only in apples sampled from the east side of the trees. In addition, fruit grown on the east side of the trees (with north–south planting direction) obtain morning sunlight, which is assumed to be less intense than sun radiation in the evening, the latter affecting directly apples grown on the west side of the trees. In 2023, only PD (p = 0.021) and P+PD (p = 0.001) led to significantly higher anthocyanin contents as compared to the UTC, irrespective of the other investigated factors (fruit side, exposition and canopy position), while P did not differ significantly from the UTC (Figure 7, right and Supporting Information Table S2). It must be noted that values for P+PD were highest for all fruit sides and exposition with respect to the other treatments. ‘Rosy Glow’ apples require a 40% red peel color coverage, higher than ‘Nicoter’, for marketing the variety under the trade name ‘Pink Lady^®’, which, in the three sampling years, was reached quantitatively by all treatments on the first harvest date (Figure 8). Although a quantitative amount of fruit with sufficient red peel color was obtained irrespective of the treatment, the anthocyanin content was enhanced following all defoliation treatments.

3.4. Crop Load

The average number of fruits per tree in control trees of ‘Nicoter’ (Figure 9A) was highest in 2022 (137 apples per tree), followed by 2023 (111 apples) and 2021 (94 apples), the latter two showing no significant differences between each other. In ‘Rosy Glow’, the number of fruits increased from 49 over 78 to 104 apples per tree in 2021, 2022 and 2023, respectively (Figure 9B), with significant differences only between 2021 and 2023. Differences in crop load over several years, in particular, regarding consecutive harvests, are within the expected fluctuations and can be influenced by alternate bearing, weather conditions or fruit thinning. To contain these variations, the same trees were used for our experiments over the three harvest years.

4. Discussion

The discussion is divided into three main parts. In the first part, we discuss the correlation of the color index with the anthocyanin contents. In the second part, we attempt to decipher the reasons for different red color formation among the three years, while in the third part, the effects of defoliation treatments on anthocyanin formation are discussed.

4.1. Correlation of the Color Index with the Anthocyanin Contents in Apple Peels

CI and anthocyanins in apple peels of ‘Nicoter’ and ‘Rosy Glow’ correlated linearly for both varieties, with coefficients of determination R² > 0.95. It is well-known that sunlight affects the parameters defining the color index [16], namely Δa, referring to the ratio between red and green color, Δb, the relationship between yellow and blue color, and ΔL, the lightness-to-darkness ratio. Our experiments showed that CI and anthocyanin contents in apple peels formed distinct groups, depending on the fruit side and on the canopy position. Thus, the sun-exposed fruit side exhibited clearly higher values of both parameters than the sun-averted one for ‘Nicoter’ in 2021, 2022 and 2023 and for ‘Rosy Glow’ in 2023, while the difference in the latter variety was less obvious in 2021 and 2022. Likewise, CI and anthocyanins on the sun-exposed fruit side could be distinguished in fruit from outside the tree canopy, showing elevated values, and fruit harvested from inside the canopy with comparably lower values. Despite the differences between these groups, CI and anthocyanins correlate well throughout all samples, a fact that had been shown also by the group of Park [11,17]. For this reason, the following parts of the discussion will be limited to deciphering the influence of crop load, meteorological conditions and defoliation treatments only on the anthocyanin contents.

4.2. Effect of Crop Load and Meteorological Conditions on the Anthocyanin Contents

The anthocyanin contents in apple peels of control (UTC) ‘Nicoter’ and ‘Rosy Glow’ trees varied among the sampling years and were highest in 2022 (averaged 9.62 mg 100 g⁻¹ fresh weight) and lowest in 2021 (6.00 mg 100 g⁻¹ fresh weight) for ‘Nicoter’ (Figure 9A), while they decreased from 2022 over 2021 to 2023 in ‘Rosy Glow’ (Figure 9B, from 11.57 over 10.50 to 8.83 mg 100 g⁻¹ fresh weight). At first, we supposed that the crop load affected red color formation of the control trees in the three years, assuming that lower crop loads resulted in elevated red color formation [18,19,20]. Unexpectedly, we found a similar behavior between the crop load (Figure 9) and the anthocyanin contents (Figure 5) for ‘Nicoter’. In ‘Rosy Glow’, we did not observe any trend between anthocyanin contents and crop load either: the number of fruits slightly increased, in fact, from 2021 to 2022 (Figure 9B), without any effect on the anthocyanin content (Figure 7). We therefore concluded that the crop load did not have any major effect on the fruit color in the present experiment.

The climatic conditions before harvest (see Figure 2 and Figure 3) varied in the three sampling years and might explain the elevated amounts of anthocyanins in apple peels of the control experiment in 2022 in comparison to 2021 and 2023. Light incidence, especially in the last two weeks before harvest, is crucial for fruit anthocyanin formation [2], as has been shown for several apple varieties [15]. The wavelength of incoming light determines the type and extent of metabolite formation in fruit trees [21], such as the synthesis of chlorophyll, carotenoids, and anthocyanins. In a previous work [8], we have shown the influence of P, PD and P+PD on the extent of photosynthetically active radiation (PAR, wavelengths between 400 and 700 nm) in different positions of the tree canopy. Besides PAR, light in the ultraviolet (UV) region determines the synthesis of specific metabolites and especially the accumulation of anthocyanins [2]. Since the overall apple peel color is determined by a combination of various metabolites, the global radiation, referring to incoming light of the entire wavelength range, provides information about the impact of sunlight in the orchard. The cumulative global radiation in the last 14 days before harvest could explain at least in part the lower anthocyanin content in the UTC of ‘Rosy Glow’ apple peels in 2023 compared to the other years. While, in 2023, the cumulative global radiation in the last 14 days before harvest of ‘Rosy Glow’ amounted to 10.8 MJ m⁻², it was higher in 2021 and 2022 (15.3 MJ m² and 12.6 MJ m⁻², respectively, see Supporting Information Figure S2). The higher values in 2021 and 2022 might account for the higher anthocyanin contents in ‘Rosy Glow’ apple peels with respect to 2023, since anthocyanin formation is triggered by sun radiation. In ‘Nicoter’, the sum of global daily radiation was not indicative of changes in anthocyanins of the UTC (Supporting Information Figure S1): in 2022 (24.7 MJ m⁻²), with the highest anthocyanin contents, the sum of global radiation in the 14 days before harvest was slightly lower than in 2021 and 2023 (24.8 MJ m⁻² and 28.2 MJ m⁻², respectively). Merely four days before the ‘Nicoter’ harvest, the global daily radiation in 2022 (sum 8.7 MJ m⁻²) rose above the values in the other years (2021: 6.2 MJ m⁻²; 2023: 8.2 MJ m⁻²). No correlation between the cumulative global radiation and the anthocyanin contents during the three harvest years of ‘Nicoter’ could thus be observed. Global radiation before harvest of ‘Nicoter’ was generally higher than before the harvest of ‘Rosy Glow’, facilitating anthocyanin production in the former variety. It must be noted that UV rather than global radiation influences anthocyanin formation, and that other metabolites, such as chlorophyll and carotenoids, further play a vital role in apple peel color formation. A profound study of the correlation between different wavelengths of incoming light and the accumulation of specific metabolites following defoliation treatments might thus provide further useful information.

Besides solar radiation, temperature might have played a role in explaining the differences in anthocyanin contents of the UTC apple peels among the years. According to literature, anthocyanin biosynthesis is stimulated by low temperatures [22,23,24]. In fact, minimum average temperatures in the 14 days before harvest in ‘Rosy Glow’ (Figure 3) were lower in 2021 (1.4 °C) than in 2022 (8.7 °C) and 2023 (9.0 °C), which fits with the inferior anthocyanin contents of UTC apple peels in 2023, but did not result in considerable anthocyanin differences between 2021 and 2022 (Figure 7 and Figure 9B). In ‘Nicoter’, the average minimum temperatures during the 14 days before harvest were rather similar (12.2, 14.1 and 14.4 °C in 2021, 2022 and 2023, respectively, Figure 3A), while the anthocyanin contents in the UTC decreased from 2022 over 2023 to 2021 (Figure 5 and Figure 9A). However, a decrease in minimum temperatures down to 7.5 °C was observed in 2022—the year with the highest anthocyanin contents—a few days before harvest, while the temperature minima in 2021 in 2023 continuously rose up to 15.5 and 16.8 °C, respectively, in the last week before harvest. As reported by Lin-Wang et al., “a single night of low temperature is sufficient to induce anthocyanin biosynthesis” [25]. The requirement for anthocyanin synthesis of only one night with low temperatures stands in agreement with our findings that the UTC of ‘Nicoter’ apple peels exhibited the highest anthocyanin values in 2022, when minimum temperatures were considerably lower in the last three days prior to harvest (as low as 7.5 °C) compared to 2021 (minimum 11.7 °C) and 2023 (14.1 °C). The authors further suggest that anthocyanins are reduced if the temperature stays above 20 °C. In our case, maximum temperatures in the last 14 days before harvest reached between 21.8 and 38.7 °C in ‘Nicoter’ and 14.3 and 28.8 °C in ‘Rosy Glow’. However, temperatures constantly dropped to their minimum at night (the lowest values were obtained at ca. 6 a.m. and 7 a.m. for ‘Nicoter’ and ‘Rosy Glow’). Minima did not exceed 16.0, 17.9 and 20.0 °C for ‘Nicoter’ and 14.4, 12.0 and 13.3 °C for ‘Rosy Glow’ in 2021, 2022 and 2023, respectively. Although low temperatures are known to promote anthocyanin production, while high temperatures inhibit their synthesis, a complex interplay of sunlight intensities and temperatures renders an exact prediction of necessary temperature values and fluctuations for optimum anthocyanin formation impossible [26,27,28]. It must be noted that temperatures before and at harvest were considerably higher for ‘Nicoter’ (on average, 30.1, 29.4 and 31.5 °C in 2021, 2022 and 2023, respectively), which were harvested in the middle of September, than for ‘Rosy Glow’ (19.4, 22.5 and 20.3 °C in 2021, 2022 and 2023, respectively) harvested at the beginning of November. Nevertheless, genetic factors play a crucial role in anthocyanin formation, rendering a direct comparison of the two varieties problematic.

Consequently, the trend in global solar radiation and the temperature minima during the three sampling years stands in agreement with the anthocyanin contents measured in peels of the control experiment for ‘Rosy Glow’ apples, while temperature maxima could not account for any anthocyanin changes in the UTC of this apple variety. In ‘Nicoter’, on the other hand, lower temperature maxima and minima correlate with elevated peel anthocyanin contents in the UTC of 2022.

4.3. Effect of Defoliation Treatments on Anthocyanin Contents

Since annual variations in the weather conditions during different harvest years can prevent the presence of adequate levels of anthocyanins in apples—especially when a specific amount of red fruit color is a prerequisite for marketing the product—novel techniques for improving apple fruit coloration have emerged. Leaf removal in the pre-harvest period has been shown to increase the coloration of apples by enhancing the light penetration in the tree canopy [7,8]. Pneumatic defoliation has the advantage to mechanize leaf removal and reduce the costs of manual leaf removal, which is often adopted in some Asian countries [29]. Evidence of the effects of pneumatic defoliation is reported in our previous study and by Win et al., who applied PD to ‘Picnic’ apple trees [8,17]. These latter findings are in line with our study, showing good correlation between the color index (CI) and the anthocyanin contents in ‘Nicoter’ and ‘Rosy Glow’ apple peels. Although several articles confirm the positive influence of pruning and defoliation treatments on apple fruit coloration, only a few reports investigate their effects comparing different harvest years, thus neglecting the potential of annual variations in light regime and temperature affecting anthocyanin synthesis. Matsumoto et al. stated that defoliation enhances the fruit skin color up to 1.7-fold in comparison to apples from non-defoliated trees especially in years where poor coloration is observed in general [30]. The favorable temperature conditions in 2022 for anthocyanin production before the harvest of ‘Nicoter’ in general and the consequent high anthocyanin contents in the UTC (on average 9.62 mg 100 g⁻¹ fresh weight) suggest that a further increase in the values by P, PD and P+PD in our experiments was only possible for ‘Nicoter’ in years where deficient coloration was observed. Thus, values increased on average by 27% from 6.00 mg to 7.63 mg 100 g⁻¹ fresh weight in 2021 and by 24% from 7.45 mg to 9.21 mg 100 g⁻¹ fresh weight in 2023. In 2022, they enhanced only by 7% to 10.26 mg 100 g⁻¹ fresh weight, in line with the results of Matsumoto et al. [30]. In ‘Rosy Glow’ apple peels, anthocyanin contents increased by 62% (from 10.24 mg to 16.60 mg 100 g⁻¹ fresh weight) only in fruit grown on the east side of the tree upon leaf removal in 2022 (merely 16% on the west side). This might be attributed to the lower anthocyanin content in the UTC on the east with respect to the west side, the former obtaining the less intense morning sunlight. Enhancements in the other sampling years were independent of the fruit exposition to the east or west side. In 2021, anthocyanins increased by 70% (10.50 mg to 17.80 mg 100 g⁻¹ fresh weight), and in 2023, by 23% (8.83 mg to 10.87 mg 100 g⁻¹ fresh weight). We further elaborated that a clear distinction between the fruit sides (sun-averted and sun-exposed) and the canopy position of the apple is necessary for a better understanding of the effects of the defoliation on the fruit color. In fact, a three-way interaction (p < 0.001) of the treatment with the canopy position (inside, outside) and the fruit side (sun-exposed, sun-averted) was observed for ‘Nicoter’ in 2021, while a two-way interaction (p = 0.031) of the treatments with the fruit side was noted in 2023. We explain such effect by considering that only the anthocyanin contents on the sun-exposed fruit side were enhanced upon all treatments P, PD and P+PD, while no increase was evidenced on the sun-averted fruit side for any of the treatments.

5. Conclusions

Our study showed that the CI correlates well with the anthocyanin contents in apple peels of ‘Nicoter’ and ‘Rosy Glow’. Peel anthocyanins can be enhanced by defoliation treatments, but the influence of P, PD and P+PD depended on the sampling year. An increase in anthocyanins was observed particularly in years with generally poor anthocyanin formation such as 2022. Defoliation treatments enhanced the anthocyanin contents of ‘Rosy Glow’ apple peels in all three sampling years (2023 only PD and P+PD), but in 2022, only in fruit grown on the east side of the trees. An increase in ‘Nicoter’ was observed in 2023 only for the sun-exposed fruit side, and in 2021, for the sun-exposed side of fruit grown inside the tree canopy, as well as the sun-averted side of fruit from outside the canopy (only for PD and P+PD). The latter involves fruit positions that obtain a medium amount of sunlight, in comparison to highly exposed and very shaded ones. In most combinations of year and variety, P, PD and P+PD were equally efficient in enhancing peel anthocyanin contents. Despite the increase in anthocyanins for both varieties, only in ‘Nicoter’ we found that the percentage of apples with >33% red peel color, prerequisite for being sold under the trade name Kanzi^®, was elevated upon defoliation ‘Rosy Glow’ apples already obtained the marketing goal of >40% red peel color for Pink Lady^® apples in the untreated control. Given the good overall results regarding the improvement of CI and anthocyanin values in both varieties, a defoliation treatment might be recommendable for bicolored apple varieties to enhance fruit peel color. From our results, it seems the combination of P+PD is not advantageous over the single treatments of P or PD.

Temperature minima in the pre-harvest period of ‘Nicoter’ were higher in 2021 and 2023 with respect to 2022, possibly leading to elevated anthocyanin contents in the control experiment in the latter year and cancelling any effect of the treatments. This means that in years with unfavorable conditions for anthocyanin formation, such as elevated daily temperature minima or low solar radiation, defoliation treatments may be advisable for bicolored apples to obtain the desired red peel coloration. Our study in fact confirms that anthocyanin contents are significantly enhanced in ‘Nicoter’ and ‘Rosy Glow’ apple peels, especially upon PD and P+PD. Nevertheless, apple peel coloration involves a complex interplay of different pigments, whose formation is favored under different environmental conditions (light of specific wavelengths, temperatures). Future research should therefore be directed towards a thorough understanding of the influence of defoliation on molecular interactions generating apple peel color.