Methods common to all experiments

Experiments were conducted at the University of Guelph, Horticultural Experiment Station, Simcoe, ON (42°51′40″ N, 80°16'8″ W). Trees were supplied with trickle irrigation via 2 L h⁻¹ pressure-compensating emitters spaced 45 cm apart and watered daily during the growing season with an equivalent of ∼2.5 cm of water weekly (adjusted for natural rainfall) on a schedule of six irrigation run-times per day every 4 h (20 min per event). Standard cultural and pest management practices for Ontario were followed (Ontario Ministry of Agriculture, Food and Rural Affairs 2018). Granular fertilizer (N-P-K) was applied in the spring each season prior to bud-break, and rates were based on leaf analyses recorded the prior season. Weeds were controlled within a 1-m strip on each side of the tree row using 1% (v/v) glyphosate applications made mid-May, June, and July. A permanent sod culture was established at the time of planting in the row middle using a mixture of 40% perennial rye and 60% red fescue (Vineland Growers, Vineland, ON). Daily maximum and minimum air temperatures, precipitation, and solar radiation were recorded by a weather station located within 500 m of the research orchards. Tree trunk circumference at 30 cm above the graft union was measured at the beginning and end of each growing season, from which trunk cross-sectional area (TCSA) was calculated. All treatment trees were selected for adequate and uniform bloom. Return bloom was estimated by counting the number of flowering and non-flowering spurs on four limbs per tree. Between 2017 and 2019 leaf phytotoxicity and leaf drop were assessed using a six-point phytotoxicity scale (0 = none; 1 = <10% very slight; 2 = 11%–20% or slight; 3 = 21%–50% or moderate; 4 = 51%–75% high; 5 = ≥75% very high). For 2020, leaf phytotoxicity was assessed using a nine-point scale assessed using the Brevis Rating (1 = no leaf necrosis; 2 = light yellow discoloration between the veins; 3 = yellow discoloration between the veins, beginning of necrosis on the edge and on the tip of the leaf; 4 = strong yellow discoloration between the veins and beginning of necrosis on the edge of the leaf (1–2 mm) from the tip… 9 = leaf entirely necrosed and curled, sometimes a small green area may persist around the central vein and near the stem) (scale provided by Adama Canada, Winnipeg, MB). At harvest, the total yield and number of fruit per tree were recorded. Mean fruit weight was calculated based on the total yield divided by the total number of fruit per tree.

Experiment 1 — Royal Gala/B.9 (2017)

This experiment was conducted in 2017 on a mature block of Royal Gala on B.9 rootstock planted in 2002 at a spacing of 2.5 m × 4.5 m (889 trees/ha) and trained to a vertical axe system. The orchard soil consisted predominately of a Watford fine sandy loam (Brunisolic Grey Brown Luvisol) (Presant and Acton 1984) with good drainage and soil textures consisting mainly of glaciolacustrine sands modified by wind sorting over fine sand and loamy fine sand at depths greater than 50 cm (Hohner and Presant 1989). Treatments were applied using a commercial air-blast sprayer (Model Turbomist 30 P, Slimline Manufacturing, Penticton, BC, Canada) to single tree plots, travelling at 3.1 km h⁻¹, 1380 kPa, 1020 L ha⁻¹ which equated to tree row volume (TRV) dilute (Sutton and Unrath 1988). The sprayer tower boom was equipped with 12 air-induction nozzles (TRX80-VK, TeeJet Technologies, Louisville, KY, USA) per boom (side); however, the axial fan was turned off to prevent spray drift to the adjacent row. In addition, to minimize treatment interference caused by spray drift, experimental units were separated by at least one guard tree.

The experimental design comprised a random complete block with seven replications and nine treatments. Treatments consisted of: (i) an untreated control; (ii) a hand-thinned control; (iii) a tank mix of 1500 mg L⁻¹ carbaryl (Sevin XLR Plus, Bayer CropScience, Guelph, ON, Canada) and 75 mg L⁻¹ 6-BA (Maxcel, Valent BioSciences, Libertyville, IL, USA); (iv)–(ix) 175 or 350 mg L⁻¹ MET (Brevis 46701; Adama Canada, Winnipeg, MB, Canada) applied at 8–10 mm, 15 mm, or 20 mm fruitlet diameter. Treatment (iii) and the 8–10 mm fruitlet diameter MET treatments were applied on 30 May 2017 (king = 8.8 mm, n = 26; lateral = 8.2 mm, n = 26). MET treatments targeting a fruitlet diameter of 15 mm were applied on 7 June 2017 (king = 17.3 mm, n = 26; lateral = 16.0 mm, n = 26) and the 20 mm treatments were applied on 12 June 2017 (king = 22.3 mm, n = 26; lateral = 20.2 mm, n = 26). All spray treatments included 0.125% LI 700 non-ionic spray adjuvant (Loveland Products Canada Inc., Dorchester, ON, Canada). After natural fruit drop, the hand-thinned control was thinned on 20 June 2017 by singling fruit and spacing them approximately 10 cm apart. The date of full bloom of was 15 May 2017.

Horticultural measurements

Four scaffold branches—two on the east and two on the west side of the tree—were selected prior to bloom to determine fruit set. On 11 May 2017, the number of flower clusters per branch were counted on each marked limb. The number of fruit set per limb were counted again on 7–8 Sept. 2017. These data were averaged and used to calculate percent fruit set (number of fruit set divided by number of flowers). Leaf phytotoxicity in plots treated at a fruitlet diameter of 8–10 mm with 350 mg L⁻¹ MET was assessed on 8 and 15 June and 20 July 2017. Fruit were harvested on 18–20 Sept. 2017 and a sample of approximately 100 fruits (∼20 kg) was placed in cold storage (∼2 °C) for subsequent grading on a commercial colour sorting grading line on 12 and 15 Jan. 2018. All fruit from each sample were graded using a commercial sorting line (MAF RODA, MAF Agrobotic, Montauban Cedex, France) which relies on cameras and sensors to weigh and size each individual fruit. Individual fruit weights, sizes, and colour were recorded. Fruit length and diameter (L:D) ratios were also determined by image analyses from the grading line. Fruit were then separated according to their weight into 10 size categories expressed as an average count size category, which was the number of apples needed to fill a 20 kg box (Canadian Food Inspection Agency 2022). The total weight of fruit per tree was calculated for each count size category: (1) <96 g (count size 216); (2) 96–108 g (count size 198); (3) 109–116 g (count size 175); (4) 117–126 g (count size 163); (5) 127–137 g (count size 150); (6) 138–151 g (count size 138); (7) 152–168 g (count size 125); (8) 169–190 g (count size 113); (9) 191–215 g (count size 100); (10) 216–237 g (count size 88); (11) 238–264 g (count size 80); (12) 265–297 g (count size 72); (13) 298–339 g (count size 64); (14) 340–396 g (count size 56); and (15) ≥397 g (count size 48). Return bloom was measured in the spring of 2018 by measuring the ratio (expressed as a percent) of flowering and non-flowering spurs on four limbs per tree.

Experiment 2 — Honeycrisp/B.9 (2017)

This experiment was conducted in 2017 on a mature block of Honeycrisp on B.9 rootstock planted in 2005 at a spacing of 2 m × 4.5 m (1111 trees/ha) and trained to a vertical axe system. The orchard soil consisted predominately of Oakland loamy fine sand (Presant and Acton 1984) with imperfect drainage and soil textures consisting mainly of 40–100 cm sandy sediments over gravelly sandy till (Hohner and Presant 1989). Treatments were applied using a commercial air-blast sprayer (Model Turbomist 30 P, Slimline Manufacturing, Penticton, BC, Canada) to single tree plots, travelling at 3.2 km hr⁻¹, 1380 kPa, 1246 L ha⁻¹ which equated to TRV dilute (Sutton and Unrath 1988). The sprayer tower boom was equipped with 12 air-induction nozzles (TRX80-VK, TeeJet Technologies, Louisville, KY, USA) per boom (side); however, the axial fan was turned off to prevent spray drift to the adjacent row. In addition, to minimize treatment interference caused by spray drift, experimental units were separated by at least one guard tree.

The experimental design comprised a randomized complete block with seven replications and seven treatments. Treatments consisted of (i) a hand-thinned control; (ii) 1500 mg L⁻¹ carbaryl; (iii) 10 mg L⁻¹ NAA (Fruitone L, AMVAC Chemical Corporation, Los Angeles, CA, USA); (iv)–(vii) 175 or 350 mg L⁻¹ MET applied at 8–10 mm or 20 mm fruitlet diameter. Treatments (ii), (iii), and the 8–10 mm diameter treatments of MET were applied on 30 May 2017 (king = 10.6 mm, n = 26; lateral = 8.7 mm, n = 26). The 15–20 mm diameter treatments of MET were applied on 12 June 2017 (king = 22.5, n = 26; lateral = 20.3, n = 26). Treatment (iii) included 0.125% LI 700 non-ionic spray adjuvant (Loveland Products Canada Inc., Dorchester, ON). After natural fruit drop, the hand-thinned control was thinned on 20 June 2017 by singling fruit and spacing them approximately 10 cm apart. The date of full bloom of was 15 May 2017.

Horticultural measurements

Four scaffold branches—two on the east and two on the west side of the tree—were selected prior to bloom to determine fruit set. On 12 May 2017 the number of flower clusters per branch were counted on each marked limb. The number of fruit set per limb were counted again on 8 Sept. 2017. These data were averaged and used to calculate percent fruit set (number of fruit set divided by number of flowers). Leaf phytotoxicity in plots treated at a fruitlet diameter of 8–10 mm with 350 mg L⁻¹ MET was assessed on 8 and 15 June and 20 July 2017. Fruit were harvested on 12 Sept. 2017 and a sample of approximately 100 fruits (1 bushel) was placed in cold storage (∼2 °C) for subsequent grading on a commercial colour sorting grading line on 17 Jan. 2018 as described for Experiment 1.

Experiment 3 — Ambrosia/M.9 2017

This experiment was conducted in 2017 on a mature block of Ambrosia on M.9 rootstock planted in 2012 at a spacing of 0.68 m × 4.0 m (3676 trees/ha) and trained to a slender spindle system. The orchard soil consisted predominately of Wilsonville sandy loam (Presant and Acton 1984) with rapid drainage and soil textures consisting of 40 cm of sandy loam over gravelly sandy till (Hohner and Presant 1989). Treatments were applied using a commercial air blast sprayer (GB Irrorazione Diserbo, Model Laser P7, Italy) to two-tree plots at 1379 kPa, 279 L ha⁻¹, which equated to TRV pesticide dilute (Sutton and Unrath 1988). The sprayer was equipped with 5=five nozzles per boom (side) and large axial fan to move the spray into the canopy. To minimize treatment interference caused by spray drift, experimental units were separated by at least one guard tree.

The experimental design comprised a randomized complete block with seven replications and seven treatments. Treatments consisted of (i) an untreated control; (ii) a hand-thinned control; (iii) 1500 mg L⁻¹ carbaryl (Sevin XLR Plus); (iv) 150 mg L⁻¹ 6-BA; (v)–(x) 175 or 350 mg L⁻¹ MET applied at 8–10 mm, 15 mm, or 20 mm fruitlet diameter. Treatments (iii), (iv), and the 8–10 mm diameter treatments of MET were applied on 30 May 2017 (king = 7.6 mm, n = 26; lateral = 6.6 mm, n = 26). The 15 mm diameter treatments of MET were applied on 7 June 2017 (king = 14.6, n = 26; lateral = 12.6, n = 26). The 20 mm diameter treatments of MET were applied on 12 June 2017 (king = 18.6, n = 26; lateral = 17.8, n = 26). After natural fruit drop, the hand-thinned control was thinned on 20 June 2017 by singling fruit and spacing them approximately 10 cm apart. The date of full bloom of was 15 May 2017.

Horticultural measurements

Four scaffold branches—two on the east and two on the west side of the tree—were selected prior to bloom to determine fruit set. On 10 May 2017 the number of flower clusters per branch were counted on each marked limb. The number of fruit set per limb were counted again on 22 Sept. 2017. These data were averaged and used to calculate percent fruit set (number of fruit set divided by number of flowers). Leaf phytotoxicity and leaf drop in plots treated at a fruitlet diameter of 8–10 mm with 350 mg L⁻¹ MET were assessed on 8 and 15 June and 20 July 2017. Fruit were harvested on 27 Sept. 2017 and a sample of approximately 100 fruits (20 kg) was placed in cold storage (∼2 °C) for subsequent grading on a commercial colour sorting grading line on 15 Jan. 2018 as described for Experiment 1.

Experiment 4 — Royal Gala/B.9 (2018)

This experiment was conducted in 2018 on a mature block of Royal Gala on B.9 rootstock planted in 2002 at a spacing of 2.5 m × 4.5 m (889 trees/ha) and trained to a vertical axe system. The orchard soil consisted predominately of a Wattford fine sandy loam (Presant and Acton 1984) with good drainage and soil textures consisting mainly of glaciolacustrine sands modified by wind sorting over fine sand and loamy fine sand at depths greater than 50 cm (Hohner and Presant 1989). Treatments were applied using a commercial air blast sprayer (GB Irrorazione Diserbo) at 1379 kPa, 947 L ha⁻¹, which equated to TRV pesticide dilute (Sutton and Unrath 1988). The sprayer was equipped with five nozzles per boom (side) and large axial fan to move the spray into the canopy. To minimize treatment interference caused by spray drift, experimental units were separated by at least one guard tree. Applications were made in low wind conditions, between 0630–1030.

The experimental design comprised a random complete block with seven replications and eight treatments. Treatments consisted of (i) an untreated control; (ii) a hand-thinned control; (iii) a tank mix of 1500 mg L⁻¹ carbaryl (Sevin XLR Plus) and 75 mg L⁻¹ 6-BA; (iv) 165 mg L⁻¹ MET (Brevis 46701), (v) 248 mg L⁻¹ MET; (vi) 330 mg L⁻¹ MET; (vii) 413 mg L⁻¹ MET; and (viii) 480 mg L⁻¹ MET. Treatment (iii) was applied on 30 May 2018 (king = 8.7 mm, n = 50; lateral = 7.3 mm, n = 50); and treatments (iv)–(viii) were applied on 2 June 2018 (king = 11.5 mm, n = 50; lateral = 9.9 mm, n = 50). All spray treatments included 0.05% Regulaid non-ionic spray adjuvant (2-butoxyethanol, poloxalene, monopropylene glycol, KALO, Inc., Overland Park, USA). After natural fruit drop, the hand-thinned control was thinned on 21 June 2018 by singling fruit and spacing them approximately 10 cm apart. The date of full bloom of was 20 May 2018.

Horticultural measurements

Four scaffold branches—two on the east and two on the west side of the tree—were selected prior to bloom to determine fruit set. On 17 May 2018, the number of flower clusters per branch were counted on each marked limb. The number of fruit set per limb were counted again on 21 June 2018. These data were averaged and used to calculate percent fruit set (number of fruit set divided by number of flowers). Leaf phytotoxicity and leaf drop were assessed in plots treated with 2.75 L ha⁻¹ and 3.2 L ha⁻¹ MET on 4, 12, and 19 June 2018. Fruit were harvested on 18–20 Sept. 2018, and sample of approximately 100 fruits (20 kg) was placed in cold storage (∼2 °C) for subsequent grading on a commercial colour sorting grading line on 5 Mar. 2019 as described for Experiment 1.

Experiment 5 — Honeycrisp/M.26 (2018)

This experiment was conducted in 2018 on a mature block of Honeycrisp on M.26 planted in 2008 at a spacing of 1.2 m × 4.0 m (2083 trees/ha) and trained to a super spindle system. The orchard soil consisted predominately of well-drained Wattford fine sandy loam (Presant and Acton 1984) with soil textures consisting mainly of glaciolacustrine sands modified by wind sorting over fine sand and loamy fine sand at depths greater than 50 cm (Hohner and Presant 1989). Treatments were applied using a commercial air blast sprayer (GB Irrorazione Diserbo) at 1379 kPa, 917 L ha⁻¹, which equated to TRV pesticide dilute (Sutton and Unrath 1988). The sprayer was equipped with five nozzles per boom (side) and large axial fan to move the spray into the canopy. To minimize treatment interference caused by spray drift, experimental units were separated by at least one guard tree. Applications were made in low wind conditions, between 0630–1030.

The experimental design comprised a randomized complete block with seven replications and nine treatments. Treatments consisted of (i) an untreated control; (ii) a hand-thinned control; (iii) 1500 mg L⁻¹ carbaryl (Sevin XLR Plus); (iv) 10 mg L⁻¹ NAA (Fruitone L); (v) a tank mix of 1500 mg L⁻¹ carbaryl and 10 mg L⁻¹ NAA; (vi) 165 mg L⁻¹ MET (Brevis 46701); (vii) 248 mg L⁻¹ MET; (viii) 330 mg L⁻¹ MET; and (ix) 413 mg L⁻¹ MET. Treatments (iii)–(v) were applied on 30 May 2018 (king = 9.2 mm, n = 50; lateral = 6.7 mm, n = 50) and treatments (vi)–(ix) were applied on 2 June 2018 (king = 12.0 mm, n = 50; lateral = 9.8 mm, n = 50). All spray treatments included 0.05% Regulaid non-ionic spray adjuvant. After natural fruit drop, the hand-thinned control was thinned on 18 June 2018 by singling fruit and spacing them approximately 10 cm apart. The date of full bloom of was 20 May 2018.

Horticultural measurements

Four scaffold branches—two on the east and two on the west side of the tree—were selected prior to bloom to determine fruit set. On 18 May 2018, the number of flower clusters per branch were counted on each marked limb. The number of fruit set per limb were counted again on 18 June 2018. These data were averaged and used to calculate percent fruit set (number of fruit set divided by number of flowers). In addition, the number of fruit set per flowering spur on the marked limbs was recorded.

Leaf phytotoxicity and leaf drop in plots treated with 2.2 and 2.75 L ha⁻¹ MET were assessed on 4, 12, and 19 June 2018. Fruit were harvested on 24 Sept. 2018, and a sample of approximately 100 fruits (20 kg) was placed in cold storage (∼2 °C) for subsequent grading on a commercial colour sorting grading line on 27 Feb. 2019 as described for Experiment 1.

Experiment 6 — Ambrosia/M.9 (2019)

This experiment was conducted in 2019 on a mature block of Ambrosia on M.9 rootstock planted in 2012 at a spacing of 0.68 m × 4.0 m (3676 trees/ha) and trained to a super spindle system. The orchard soil consisted predominately of Wilsonville sandy loam (Presant and Acton 1984) with rapid drainage and soil textures consisting of 40 cm of sandy loam over gravelly sandy till (Hohner and Presant 1989). Treatments were applied using a commercial air blast sprayer (GB Irrorazione Diserbo) at 1379 kPa, 620 L ha⁻¹, which equated to TRV pesticide dilute (Sutton and Unrath 1988). The sprayer was equipped with five nozzles per boom (side) and large axial fan to move the spray into the canopy. To minimize treatment interference caused by spray drift, experimental units were separated by at least one guard tree. Applications were made in low wind conditions, between 0630–1030.

The experimental design comprised a randomized complete block with seven replications and 11 treatments. Treatments consisted of (i) untreated control; (ii) hand-thinned control; (iii) 1500 mg L⁻¹ carbaryl (Sevin XLR Plus); (iv) a tank mix of 1500 mg L⁻¹ carbaryl + 75 mg L⁻¹ 6-BA; (v) 175 mg L⁻¹ MET (Brevis 46701); (vi) 263 mg L⁻¹ MET; (vii) 351 mg L⁻¹ MET; (viii) 2.92 L ha⁻¹ MET; (ix)–(xi) 1.75 L ha⁻¹ MET applied at petal fall, 8–11 mm fruitlet diameter or 16–20 mm fruitlet diameter. Treatments (iii)–(viii) were applied on 9 June 2019 targeting a fruitlet diameter of 12–14 mm (king = 12.4 mm, n = 23; lateral = 9.2 mm, n = 23). Treatment (ix) was applied at petal fall on 31 May 2019, treatment (x) on 7 June 2019 (king = 9.2 mm, n = 50; lateral = 6.8 mm, n = 50), and treatment (xi) on 18 June 2019 (king = 17.7 mm, n = 50; lateral = 13.5 mm, n = 50). Treatments (iii) and (iv) included 0.05% Regulaid non-ionic spray adjuvant. After natural fruit drop, the hand-thinned control was thinned on 4 July 2019 by singling fruit and spacing them approximately 10 cm apart. The date of full bloom of was 25 May 2019.

Horticultural measurements

Four scaffold branches—two on the east and two on the west side of the tree—were selected prior to bloom to determine fruit set. On 22 May 2019, the number of flower clusters per branch were counted on each marked limb. The number of fruit set per limb were counted again on 4 July 2019. These data were averaged and used to calculate percent fruit set (number of fruit set divided by number of flowers). Leaf phytotoxicity and leaf drop were assessed on 14, 20, and 28 June 2019. Fruit were harvested on 10–11 Oct. 2019, and a sample of approximately 100 fruits (20 kg) was placed in cold storage (∼2 °C) for subsequent grading on a commercial colour sorting grading line on 25 Jan. 2020 as described for Experiment 1.

Experiment 7 — Ambrosia/B.9 (2020)

This experiment was conducted in 2020 on a mature block of Ambrosia/B.9 rootstock planted in 2005 at a spacing of 2 m × 4.5 m (1111 trees/ha) and trained using a spindle training system. The orchard soil was the same as that in Experiment 6. Treatments were applied using a commercial air blast sprayer (GB Irrorazione Diserbo) at 1379 kPa, 816 L ha⁻¹, which equated to TRV pesticide dilute (Sutton and Unrath 1988). The sprayer was equipped with five nozzles per boom (side) and large axial fan to move the spray into the canopy. To minimize treatment interference caused by spray drift, experimental units were separated by at least two guard trees. All applications were made in low wind conditions.

The experimental designed consisted of a randomized complete block with four replications and 13 treatments. Treatments were applied to single tree plots and consisted of (i) an untreated control; (ii) a hand-thinned control; (iii) 1500 mg L⁻¹ carbaryl; (iv) 1500 mg L⁻¹ carbaryl tank mixed with 75 mg L⁻¹ 6-BA; (v)–(xiii) four rates of MET: 175.5 mg L⁻¹, 263 mg L⁻¹, 351 mg L⁻¹, and 438 mg L⁻¹ applied with or without the non-ionic surfactant Agral 90 (Norac Concepts Inc, Guelph, ON) applied at a rate of 0.05% (v/v).

The MET treatments were applied on the mornings of 1 June and repeated on 9 June corresponding to a target fruitlet diameter of 6–7 mm and 12–16 mm, respectively. A single application of carbaryl and carbaryl + 6-BA treatments were applied on the morning of 9 June. Actual fruitlet diameters on 1 June were king = 5.0 mm (n = 50) and lateral = 4.2 mm (n = 50). Fruitlet diameters on 9 June were king = 12.5 mm (n = 25) and lateral = 10.1 mm (n = 25). The hand-thinned control trees were hand thinned on 8 July by removing all but one fruit per cluster and spacing fruit ∼10 cm apart. The date of full bloom of was 25 May 2020. Agral 90 surfactant at 0.05% (v/v) was included in the carbaryl alone and tank-mix of carbaryl and 6-BA treatments.

Horticultural measurements

On 11 May prior to full bloom, a total of four main scaffold branches, two on each of the east and west sides of the tree, were selected and marked to determine fruit set by counting the number of flower clusters per branch. On 22 June, after June drop, the number of fruit set per flower cluster was counted. These data were averaged and used to calculate percent fruit set (number of fruit set divided by number of flowers). Leaf phytotoxicity and leaf drop were assessed on 16 June 2020. Fruit were harvested on 13 Oct. 2020, and a sample of approximately 100 fruits (20 kg) was placed in cold storage (∼2 °C) for subsequent grading on a commercial colour sorting grading line on 7 Nov. 2020 as described for Experiment 1. Return bloom was assessed in the spring of 2021 by counting the number of spurs with and without flowers on four limbs per tree between 1.5 m and 2 m above the ground. Return bloom was expressed as the percentage of spurs with flowers.

Statistical analyses

Data were subjected to analysis of variance using the PROC GLIMMIX procedure of SAS version 9.4 (SAS Institute, Cary, NC, USA). Average fruit weight was adjusted by using crop load as a covariate (Marini et al. 2012). Comparisons of treatments lsmeans were made using a Tukey’s multiple means comparison test and statistical significance was reported at a Type I error rate of ∝ = 0.05. Single degrees of freedom orthogonal comparison were performed to evaluate the rate and timing effects of MET. A Shapiro-Wilk test was used to test the normality of residuals. Scatterplots of studentized residuals were visually observed to test the assumption that errors were not heterogeneous. In cases where there were large deviations from the assumptions, data were transformed using log- or square root-transformation prior to analysis.

Experiment 1 — Royal Gala/B.9 (2017)

Thinning treatments had a significant effect on Royal Gala fruit set (P = 0.0002), yield (P = 0.0092), total number of fruit per tree (P = 0.0366), but no effect on TCSA, percent marketable yield, adjusted average fruit weight, crop load, and return bloom the year following treatment applications (Table 1). Trees that were untreated had the highest fruit set compared to any of the other thinning treatments. Trees treated with 350 mg L⁻¹ MET applied at 15- and 20-mm fruitlet diameter had the lowest fruit set, which was similar to the hand-thinned treatment and the treatments receiving fruitlet thinners. Based on orthogonal contrasts, MET at 15 mm and 20 mm were more effective than when applied at the 8–10 mm fruitlet stage. Fruit set was similar when MET was applied at 175 or 350 mg L⁻¹. Trees that were thinned with MET at 8–10 mm had higher yields, more fruit per tree, and higher crop loads compared with the untreated control, hand-thinned, and the other MET application timings. It is unclear why MET, when applied at 8–10 mm, would have resulted in higher fruit numbers per tree and crop loads compared to the untreated trees. Fruit size distribution was not influenced by any of the thinning treatments (data not shown). There was minimal effect of MET on leaf phytotoxicity (data not shown).

Table 1.

Influence of carbaryl (CB), 6-benzyladenine (6-BA), and various rates of metamitron applied at three different fruitlet sizes on tree growth, fruit set, and yield in 2017 of Royal Gala/B.9 apples planted in 2002.

Experiment 2 — Honeycrisp/B.9 (2017)

Thinning treatments had a significant effect on the total number of Honeycrisp fruit per tree (P = 0.0131) but had no effect on TCSA, fruit set, total yield, percent marketable yield, adjusted average fruit weight, crop load, and return bloom the year following applications (Table 2). Trees that were hand-thinned had the highest fruit set compared with any of the other trees treated with chemical thinners. Based on orthogonal contrasts, trees treated with MET at both the 8–10 mm and 15–20 mm timings had lower fruit set compared with the hand-thinned trees and were both equally effective in reducing fruit set. The response to MET applied at 175 or 350 mg L⁻¹ was similar. Trees that were hand thinned had the lowest number of fruit per tree. Trees treated with 1500 mg L⁻¹ carbaryl, 10 mg L⁻¹ NAA, or any of the MET treatments at 8–10- or 15–20-mm fruitlet diameters had similar numbers of fruit per tree, with the exception of those treated with 175 mg L⁻¹ MET applied at 8–10 mm. However, based on orthogonal contrasts, the number of fruit per tree was significantly lower for the hand-thinned trees compared with MET treatments at both the 8–10 mm and 15–20 mm stages. Orthogonal contrasts indicated that MET applied at 15–20 mm were more effective in reducing crop load than when MET was applied at 8–10 mm compared with the hand-thinned controls. Overall, there was only minor treatment effects on fruit size distribution and leaf phytotoxicity (data not shown).

Table 2.

Influence of carbaryl (CB), 1-naphthaleneacetic acid (NAA), and various rates and application timings of metamitron on tree growth, fruit set, and yield in 2017 of Honeycrisp/B.9 apples planted in 2005.

Experiment 3 — Ambrosia/M.9 2017

Thinning treatments had a significant effect on Ambrosia TCSA (P = 0.0146), fruit set (P = 0.0054), total number of fruit per tree (P = 0.0008), crop load (P < 0.0001), and return bloom (P < 0.0001), but no effect on total yield, percent marketable yield, and adjusted average fruit weight (Table 3). Fruit set was similar among all treatments, except the hand-thinned control treatment, which was slightly lower, but not statistically different from the untreated, carbaryl or 6-BA grower control treatments, or most MET treatments. Based on orthogonal contrasts, fruit set in trees treated with MET at 8–10 mm and 20 mm fruitlet diameters were higher than that in the hand-thinned control trees. Trees treated with 175 mg L⁻¹ had the greatest number of fruit per tree while trees treated with 1500 mg L⁻¹ carbaryl had the lowest number of fruit per tree. All trees treated with MET had a similar number of fruit per tree as trees left untreated or hand-thinned. The untreated trees and those treated with MET at any time or concentration had similar crop loads. In contrast, the hand-thinned and carbaryl treated trees had the lowest crop loads; these, however, were similar to trees receiving a MET treatment. There was a marked treatment effect on return bloom the following year; trees left untreated had 26% of spurs with flowers whereas hand-thinned trees had 79% of spurs with flowers. Based on the ANOVA, trees receiving MET had similar return bloom values, regardless of application timing or concentration, all of which were lower than the hand-thinned trees. However, based on orthogonal contrasts, trees which received 350 mg L⁻¹ MET had a slightly higher percentage of spurs with flowers than those treated with 175 mg L⁻¹ MET. There was minimal effect of MET on leaf phytotoxicity (data not shown).

Table 3.

Influence of carbaryl (CB), 6-benzyladenine (6-BA), and two rates of metamitron applied at various fruitlet diameters on tree growth, fruit set, and yield in 2017 of Ambrosia/M.9 apples planted in 2012.

There was a significant treatment effect on fruit size distribution (Table 4). Overall, trees that were hand-thinned had the largest fruit weights, which exceeded those from the untreated control and any trees treated with MET. These effects were largely observed in the 125, 138, and 150 size categories. There was no MET concentration or timing effect on grade size distribution compared to the untreated control trees in any of the size categories between 88 and 198; however, in the 216 count size categories, the untreated control trees and those treated with 175 mg L⁻¹ MET at 20 mm had the greatest number of fruit.

Table 4.

Influence of carbaryl (CB), 6-benzyladenine (6-BA), and two rates of metamitron applied at various fruitlet diameters on the weight of fruit per count size in 2017 of Ambrosia/M.9 apples planted in 2012.

Experiment 4 — Royal Gala/B.9 (2018)

Thinning treatments had a significant effect on Royal Gala TCSA (P = 0.0009), fruit set (P < 0.0001), the percentage of flowering spurs with zero, one, two, three and four fruits, but had no effect on return bloom (Table 5). Trees that were hand-thinned had the greatest TCSA while trees that were treated with 165 mg L⁻¹ MET (P = 0.0235) had the smallest TCSA. There were no differences in TCSA between any of the trees treated with MET and the untreated trees. Fruit set decreased with increasing rate of MET in a linear fashion (P = 0.001). Trees treated with 1500 mg L⁻¹ carbaryl combined with 75 mg L⁻¹ 6-BA and those that were hand-thinned had the lowest fruit set. Trees treated with 248 mg L⁻¹ MET or higher had similar fruit set as the hand-thinned trees, while trees treated with 165 and 248 mg L⁻¹ MET had fruit set similar to the untreated trees. With increasing rates of MET, the percentage of spurs with no fruit, two fruit, three fruit, and four fruit decreased in a highly significant fashion. Trees that had the highest percentage of fruits with more than one fruit per cluster where those left untreated and those treated with 165, 248, and 330 mg L⁻¹ MET.

Table 5.

Influence of carbaryl (CB), 6-benzyladenine (6-BA), and various rates of metamitron on fruit set in 2018 of Royal Gala/B.9 apples planted in 2002.

Thinning treatments also had a highly significant effect on Royal Gala total yield (P < 0.0001), number of fruit per tree (P =< 0.0001), percent marketable fruit (P < 0.0001), adjusted average fruit weight (P < 0.0001), and crop load (P < 0.0001) (Table 6). Increasing rates of MET decreased total yield, number of fruit per tree, and crop load in a linear fashion, but increased percent marketable fruit and adjusted average fruit weight in a quadratic and linear fashion, respectively. Trees treated with 330 mg L⁻¹ MET or higher had similar yields, fruit numbers, percent marketable fruit, and crop load as the hand-thinned control trees. Trees treated with 413 mg L⁻¹ MET or higher had similar fruit weights as the hand-thinned control and the grower control treatment of 1500 mg L⁻¹ carbaryl combined with 75 mg L⁻¹ 6-BA. There was minimal effect of MET on leaf phytotoxicity (data not shown).

Table 6.

Influence of carbaryl (CB), 6-benzyladenine (6-BA), and various rates of metamitron on fruit yield in 2018 of Royal Gala/B.9 apples planted in 2002.

Thinning treatments had a significant effect on fruit size distribution in the 88 and 125 through 198 categories (Table 7). The untreated trees had the greatest number of small fruit while trees that were hand-thinned had the greatest number of large fruit.

Table 7.

Influence of carbaryl (CB), 6-benzyladenine (6-BA), and various rates of metamitron on the amount of fruit in each size category in 2018 of Royal Gala/B.9 apples planted in 2002.

Experiment 5 — Honeycrisp/M.26 (2018)

Thinning treatments had a significant effect on Honeycrisp fruit set (P < 0.0001), the percentage of flowering spurs with zero, one, two, and three fruits, and return bloom (P < 0.0001), but had no effect on TCSA (Table 8). Fruit set decreased with increasing rates of MET in a linear fashion (P = 0.001). Trees treated with 1500 mg L⁻¹ carbaryl combined with 10 mg L⁻¹ NAA had the lowest fruit set followed by trees treated with 10 mg L⁻¹ NAA or 1500 mg L⁻¹ carbaryl. Trees treated with rates of 168 mg L⁻¹ MET and higher had similar fruit set as the hand-thinned trees. All trees that were hand-thinned or treated with fruitlet thinners, including MET, had lower fruit set than trees left untreated. With increasing rates of MET, the percentage of spurs with zero fruit, one fruit, two fruit, and three fruit decreased. Trees that had the highest percentage of fruits with two fruit per cluster were those left untreated or those treated with 165, 248, and 330 mg L⁻¹ MET. Trees that were hand-thinned, left untreated, or treated with 165, 248, and 413 mg L⁻¹ MET had the greatest percentage of spurs with one fruit. Overall, return bloom was poor for all treatments, except those treated with 10 mg L⁻¹ NAA and 1500 mg L⁻¹ carbaryl. Even trees that were hand-thinned or treated with 10 mg L⁻¹ NAA combined with 1500 mg L⁻¹ carbaryl had poor return bloom (31% and 28%, respectively), albeit higher than that observed for the untreated control. Trees treated with any of the rates of MET had similar return bloom as the untreated control trees.

Table 8.

Influence of various rates and combinations of metamitron, carbaryl (CB), and naphthaleneacetic acid (NAA) on fruit set in 2018 of Honeycrisp/M.26 apples trees planted in 2008.

Thinning treatments also had a highly significant effect on Honeycrisp total yield (P < 0.0001), number of fruit per tree (P < 0.0001), percent marketable fruit (P < 0.0001), adjusted average fruit weight (P < 0.0001), and crop load (P < 0.0001) (Table 9), however most of these effects can be attributed to the hand-thinned, carbaryl, NAA, and carbaryl combined with NAA sprays as most trees treated with MET, especially at lower rates, had little or no effect compared with the untreated control trees. While MET had little effect on yield per tree, increasing rates of MET above 248 mg L⁻¹ reduced fruit set below the untreated control and to a level similar to the hand-thinned control. Overall, the percentage of marketable fruit were low (because of poor colour; data not shown), and with respect to trees receiving the MET treatments, only those treated with 413 mg L⁻¹ MET had slightly improved marketable fruit above the untreated. None of the MET treatments improved fruit weight over the untreated control treatments. Increasing rates of MET reduced crop load in a quadratic fashion, but only trees treated with 330 or 413 mg L⁻¹ MET had lower crop loads than the untreated control trees, as did the hand-thinned control, carbaryl, NAA, and carbaryl and NAA applications. There was minimal effect of MET on leaf phytotoxicity (data not shown).

Table 9.

Influence of various rates and combinations of metamitron, carbaryl (CB), and naphthaleneacetic acid (NAA), on harvest parameters in 2018 of Honeycrisp/M.26 apple trees planted in 2008.

Honeycrisp fruit were large because of relatively light crop loads. Thinning treatments had a significant effect on fruit size distribution in all size categories except the smaller 150, 163, 198, and 216 categories (Table 10). The untreated trees had the greatest number of small fruit, but only 3% of the fruit (by weight) were smaller than the 138 size category. Trees that were hand-thinned had the greatest number of large fruit where most were distributed in the 64 to 100 count size. All trees treated with MET had similar amounts of fruit in each of the size categories ranging from 56 through 125 compared with the untreated control trees.

Table 10.

Influence of various rates and combinations of metamitron, carbaryl (CB), and naphthaleneacetic acid (NAA), on the amount of fruit in each size category in 2018 of Honeycrisp/M.26 apples planted in 2008.

Experiment 6 — Ambrosia/M.9 (2019)

Thinning treatments had a significant effect on Ambrosia fruit set (P < 0.0001), the percentage of flowering spurs with zero, one, two, and three fruits, and return bloom (P < 0.0001) but had no effect on TCSA (Table 11). Fruit set decreased with increasing rate of MET in a highly significant linear fashion (P = 0.001). Trees treated with 1500 mg L⁻¹ carbaryl or 1500 mg L⁻¹ carbaryl combined 75 mg L⁻¹ 6-BA had similar fruit set as trees treated with 351 and 438 mg L⁻¹ MET. Trees that were left untreated or hand-thinned had the highest fruit set. Whereas applications of 263 mg L⁻¹ MET at 8–11 or 16–20 mm fruitlet diameter reduced fruit set more than the petal fall application, and the 12–14 mm timing was the most efficacious. Regardless of treatment, most trees had flowers set as either zero or one fruit per spur. For example, trees left untreated had only 8% fruit with more than one fruit per flower cluster. With increasing rates of MET, the percentage of spurs with no fruit increased in a linear fashion while the percentage of spurs with one fruit decreased in a linear fashion.

Table 11.

Influence of various rates, combinations, and application timings of metamitron, carbaryl (CB), and 6-benzyladenine (6-BA), on fruit set in 2019 of Ambrosia/M.9 apples planted in 2012.

Thinning treatments also had a highly significant effect on Ambrosia total yield (P < 0.0001), number of fruit per tree (P < 0.0001), percent marketable fruit (P < 0.0001), adjusted average fruit weight (P < 0.0001), and crop load (P < 0.0001) (Table 12). Increasing rates of MET had an inverse linear relationship with yield, number of fruit per tree, percent marketable fruit, average fruit weight and crop load. Applications of 263 mg L⁻¹ MET made after the petal fall timing were more effective in reducing the total number of fruit per tree and crop load. Trees spayed with 263, 351, and 438 mg L⁻¹ MET at 12–14 mmm were equally effective as 1500 mg L⁻¹ carbaryl and 1500 mg L⁻¹ carbaryl combined with 75 mg L⁻¹ 6-BA at the same timing. Percent marketable fruit increased in a linear fashion with MET, with 438 mg L⁻¹ MET treatment resulting in 92% marketable fruit. The untreated control trees had only 23% fruit that were marketable, owing largely to poor colour (data not shown) rather than small fruit size. Petal fall and 8–11 mm fruit diameter applications of 263 mg L⁻¹ MET had a similar percentage of marketable fruit as the untreated control, while applications made at 12–14 and 16–20 mm had 89% and 62% percent marketable fruit, respectively. The two grower control applications of carbaryl and carbaryl combined with 6-BA had 85% and 86% marketable fruit, respectively. In comparison with the untreated control, MET at any of the concentrations did not increase average fruit weight, nor was there any effect of MET application timing on fruit weight; however, based on orthogonal contrasts, there was a linear increase in fruit weight with increasing concentration of MET. Carbaryl and carbaryl combined with 6-BA treatments did not improve fruit weight compared with the untreated control. Despite the minor effects on fruit weight, increasing rates of MET reduced crop load in a highly linear fashion, with all rates of MET resulting in lower crop loads compared with the untreated controls. Petal fall and 8–11 mm fruit diameter applications of 263 mg L⁻¹ MET were ineffective in reducing crop loads below that of the untreated control, whereas 12–14 and 16–20 mm applications reduced crop loads similar to the hand-thinned and grower control applications of carbaryl and carbaryl combined with 6-BA. There was minimal effect of MET on leaf phytotoxicity (data not shown).

Table 12.

Influence of various rates and combinations of metamitron, carbaryl (CB), and 6-benzyladenine (6-BA) on harvest parameters in 2019 of Ambrosia/M.9 apples planted in 2012.

Thinning treatments had a significant effect on fruit size distribution in all categories ranging from 64 through 175 (Table 13) with the exception of the 100 size category. The untreated trees had the greatest number of small fruit, but only 4% of the fruit (by weight) were smaller than the 163 size category, the minimum weight of fruit sold to the fresh/retail market. Trees that were treated with 1500 mg L⁻¹ carbaryl and 1500 mg L⁻¹ carbaryl combined 75 mg L⁻¹ 6-BA had the greatest amount of large fruit where most fruit was distributed in the 56 to 88 size categories. The untreated control trees produced fruit primarily in the 100–175 size categories, which is a respectable number of marketable fruit given the relatively high crop loads from these trees. Trees treated with MET and the grower control fruitlet thinners had most of their fruit fall in two or three size categories larger.

Table 13.

Influence of various rates and combinations of metamitron, carbaryl (CB), and 6-benzyladenine (6-BA) on the weight of fruit per count size in 2019 of Ambrosia/M.9 apples planted in 2012.

Experiment 7 — Ambrosia/B.9 (2020)

There was a significant treatment effect on fruit set (P < 0.001) (Table 14). The untreated control treatment had the highest fruit set while trees treated with 438 mg L⁻¹ MET without Agral 90 had the lowest fruit set. Trees treated with 351 mg L⁻¹ or 438 mg L⁻¹ MET with Agral 90 also had significantly lower fruit set compared with the untreated and hand-thinned controls. The 175 mg L⁻¹ or 263 mg L⁻¹ L MET treatments resulted in intermediate values, mildly affected fruit set and were statistically similar to the hand-thinned control and grower standard carbaryl and carbaryl combined with 6-BA treatments. There was a significant negative linear rate response (P < 0.0001) in fruit set; that is to say fruit set decreased with increasing rates of MET. The addition of a surfactant had no statistical effect on fruit set.

Table 14.

Influence of carbaryl (CB), 6-benzyladenine (6-BA), and various rates of metamitron with or without Agral 90 surfactant on fruit set in 2020 of Ambrosia/B.9 apple trees.

There was also a significant treatment effect on the number of fruit set per flower cluster (Table 14). Treatments that had the greatest number of flowering clusters where zero fruit set per cluster were the 438 mg L⁻¹ MET without Agral 90 followed by the 351 and 438 mg L⁻¹ MET treatments with Agral 90. Having a proportion of flowering spurs with no fruit is important because these are the spurs that will flower and fruit the following season. The untreated control treatments had the greatest percentage numerically of flower clusters with one fruit, but this was similar to the hand-thinned controls and grower standard carbaryl and carbaryl combined with 6-BA treatments as shown by the means separation. Applications of 438 mg L⁻¹ MET without Agral 90 resulted in the lowest percentage of clusters with one fruit. The 175 mg L⁻¹ or 263 mg L⁻¹ rates of MET were similar to the hand-thinned control and grower standard treatments. The untreated control trees had the highest percent of clusters with two fruit per cluster while the 438 mg L⁻¹ MET without Agral 90 and 351 to 438 mg L⁻¹ MET with Agral 90 treatments had the lowest percent of clusters with two fruit. The untreated control also had the highest percentage of flowering clusters with three fruit. Few flowering clusters had four fruit per cluster and consequently there was not treatment effect. The percentage of clusters with zero fruit increased while the percentage of clusters with one to three fruits decreased in a linear fashion with increasing rates of MET (P < 0.0001). The addition of a surfactant did not affect the number of fruit per cluster.

The thinning treatments did not significantly affect total yield per tree, total number of fruits per tree, fruit weight, or crop load (Table 15). There was a significant treatment effect on percent marketable fruit, which was highest in trees treated with 438 mg L⁻¹ MET without Agral 90. Although the ANOVA generally showed no significant treatment effects at P = 0.05, the orthogonal contrasts of the various MET rates indicated that total yield, total number of fruit, and crop load decreased, while percent marketable fruit increased, in a linear fashion with increasing rates of MET. There was a significant linear and quadratic effect of rate of MET without Agral 90 on the percentage of flowering spurs and a significant linear effect of rate of MET with Agral 90 on percentage of flower spurs. That is to say that as the rate of MET increased, the percentage of flowering spurs increased in a curvilinear (quadratic) fashion when Agral 90 was excluded, and decreased in a linear fashion when Agral 90 was included.

Table 15.

Influence of carbaryl (CB), 6-benzyladenine (6-BA), and various rates of metamitron with or without Agral 90 surfactant on harvest parameters of Ambrosia/B.9 apple in 2020.

Overall, thinning treatments did not have a marked effect on the size distribution of fruit (Table 16), with the exception of the 125, 163, and 175 size categories, which had significant treatment effects according to the ANOVA but no treatment differences shown by the lsmeans separation. Within these size categories, increasing rates of MET generally had an inverse linear effect on the amount of fruit within each category. The addition of a surfactant with MET had no measurable effect on fruit size distribution for any of the 15 size categories.

Table 16.

Influence of carbaryl (CB), 6-benzyladenine (6-BA), and various rates of metamitron with or without Agral 90 surfactant on the weight of fruit per count size of Ambrosia/B.9 apples in 2020.

Severity of leaf phytotoxicity, as measured by the Brevis rating scale, on the Ambrosia trees was low (<4 rating) but was significantly affected by the MET treatments (Table 17). Phytotoxicity was most severe in trees treated with 438 mg L⁻¹ MET without Agral 90. The 263 and 438 mg L⁻¹ rates of MET with Agral 90 also resulted in significantly greater phytotoxicity compared with the untreated control, as shown by the lsmean separation. The percent of leaves affected was very low (<2%) but was significantly higher following applications of 438 mg L⁻¹ MET without Agral 90 and 263 and 438 mg L⁻¹ rates of MET with Agral 90 compared with the untreated control. Albeit low, the orthogonal contrasts indicated that the severity and incidence of phytotoxicity increased in a linear fashion with increasing rates of MET. The addition of Agral 90 surfactant did not affect the severity of phytotoxicity but had a mild effect on the incidence of phytotoxicity, which increased on average from 1.1% to 1.4% (P = 0.05). Overall, this would have an inconsequential effect on the orchard commercially.

Table 17.

Severity of phytotoxicity following applications of carbaryl (CB), 6-benzyladenine (6-BA), and various rates of metamitron with or without Agral 90 surfactant in 2020 on Ambrosia/B.9 apple trees.

Effect of fruit development — fruitlet diameter

In three of the four experiments where we investigated different times of application, applications between 8 and 20 mm fruitlet diameter were most effective, while petal fall applications, which represents fruitlets approximately 5 to 6 mm diameter, were least effective. Brunner (2014) found the highest abscission rates when MET was applied to Fuji and Golden Delicious around 12 mm fruitlet diameter; for Fuji specifically, MET was less effective when applied at petal fall and at 16 mm fruitlet fruit diameter, whereas for Golden Delicious good thinning with MET occurred at petal fall and at 16 mm fruit fruitlet diameter. Gabardo et al. (2017) found MET was effective when applied at petal fall to Baronesa and Fuji apple trees. Greene (2014) found MET application at 16 mm and 20 mm were effective in thinning CandyCrisp apple trees. Lafer (2010) found strong overthinning when 350 mg L⁻¹ MET was applied to Elstar at 6–8 mm one year but not the next. Furthermore, when MET was applied at 12–14 mm fruitlet diameter, it only thinned in one of three years. The authors found optimal fruit size was obtained by the repeated application of MET at 6–8 mm and 12–14 mm fruit diameter. In a 3-yr study, McArtney and Obermiller (2014) observed that 350 mg L⁻¹ MET applied at petal fall to Morgan Spur Delicious did not reduce fruit set over 3 yr but found applications at 10 mm fruitlet diameter exhibited strong thinning activity. The authors concluded that yearly variation in thinning efficacy of chemical thinning sprays applied at 10 mm fruit diameter could not be explained by a thinning index based on the average daily carbon balance for 4 d beginning on the day of thinner application. In another study on Gale Gala, McArtney and Obermiller (2012) observed that 350 mg L⁻¹ MET reduced fruit set when applied at 18 mm fruitlet diameter. Radivojevic et al. (2020) observed that application of 250 mg L⁻¹ MET to Golden Delicious at 16–18 mm fruit diameter was very effective in reducing fruit set. In a Chilean study, Reginato et al (2017) found that 450 g MET ha⁻¹, applied between 4 mm (petal fall) and 28 mm fruit diameter, effectively thinned Brookfield Gala trees, even when the trees were predicted to have a strong carbohydrate balance, especially during the later application timings. Collectively, these studies indicate a wide variation in thinning response when MET is applied to fruitlets of different diameters during fruit set. This is because timing effects can be confounded by environmental factors, fruit, and tree physiological factors, and genetic differences among rootstocks and scions. There is unlikely one specific fruitlet size at which the thinning response is consistently effective.

Rate effect of MET

All seven experiments evaluated two or more MET rates at single or multiple application timings. Thinning increased with higher rates in four of the seven studies, all of which occurred in 2018, 2019, and 2020 but not in 2017. In all instances, there was a linear reduction in crop load, number of fruit per tree, and fruit set with increasing rates of MET. In Honeycrisp (2018), rates above 248 mg L⁻¹ MET were required to mount an effective thinning response. For Ambrosia (2019) 175 mg L⁻¹ MET was effective in reducing fruit set and crop load, but rates at or above 263 mg L⁻¹ MET were required to reduce the total number of fruit per tree. For Gala (2018), 248 mg L⁻¹ MET reduced fruit set and number of fruit per tree, but 413 mg L⁻¹ MET was required to reduce crop loads below the untreated control. The higher concentration of MET needed to thin Gala may be an indication of its greater difficultly to thin compared with Ambrosia and Honeycrisp (Ontario Ministry of Agriculture, Food and Rural Affairs 2021). Brunner (2014) found that MET increased fruit abscission of Golden Delicious in a concentration dependent manner and thinning occurred with MET (Brevis, 15% MET) when applied between 1–2.5 kg ha⁻¹. Lower rates of Brevis (0.825 kg ha⁻¹) did not sufficiently reduce fruit set, whereas over-thinning occurred with concentrations exceeding 2.5 kg ha⁻¹. Gabardo et al. (2017) found when MET was applied between 350 to 1750 mg L⁻¹, there were linear reductions in crop load, number of fruit per tree, and yield and a linear increase in fruit size. At high rates, the authors observed over-thinning and pre-harvest fruit drop at harvest with no improvement in return flowering the following year. As with the current study, Brunner (2014) found differences in thinning responses with the same amounts of MET applied in different years. As none of the higher rates of MET caused extensive overthinning in the present study, using increasing concentration of MET sprays, in addition to using multiple sprays in sequence, and combining MET with 6-BA or carbaryl, are all effective strategies of increase fruitlet abscission and improving thinning efficacy.

Effect on MET on fruit weight

In only one of the seven studies did MET increase adjusted fruit weight above that of the untreated control trees, even though MET increased fruit abscission and decreased crop load. In the 2018 Gala experiment, when 413 and 480 mg L⁻¹ MET was applied when king fruitlets were 11.5 mm, fruit weight was increased by 7% and 8%, respectively, but total yield per tree was reduced by 47% and 45%, respectively. In the same experiment, the hand-thinned and grower control spray of 1500 mg L⁻¹ carbaryl combined with 75 mg L⁻¹ 6-BA had higher fruit weights in comparison, but even lower crop loads. In the 2019 Ambrosia experiment, MET was very effective in reducing crop load, but crop load adjusted fruit weight was similar to the untreated control trees. These data collectively indicate that MET induced increases in fruit weight can be explained by MET’s ability to reduce crop load rather than a direct effect on fruit weight.

Effect of surfactants

Surfactants are often added to plant growth regulator sprays in increase plant uptake vis-à-vis the leaf or fruit (Stover and Greene, 2005). Greene (2014) found that when Regulaid, a non-ionic surfactant, was included in all treatments on Candy Crisp apple, extensive thinning and phytotoxicity was observed in one year. Yet when the same surfactant was included with a spray of 150 mg L⁻¹ MET the following year, there was neither increased thinning activity nor an enhancement of phytotoxicity. When McArtney and Obermiller (2012) included the organosilicone surfactant, Silwett L-77 with 350 mg L⁻¹ MET sprays to Cameo apple, it resulted in a greater reduction in Fv/Fm and fruit set. Effects on thinning were not reported. Based on limited research, further studies are required to fully evaluate the effects of surfactants when combined with MET for thinning apples.

Effect of MET on leaf phytotoxicity

In the present experiments inconsequential leaf phytotoxicity was observed when applied to Ambrosia, Gala, and Honeycrisp over the 4-yr study period. There are few reports in the scientific literature that report on leaf phytotoxicity of MET for thinning apples. Lafer (2010) reported that 350 mg L⁻¹ MET applied to Elstar apples over two seasons did not cause any leaf phytotoxicity, pygmy fruit formation, or fruit russeting. Antidotal reports from the Pacific Northwest have raised concern of higher rates of MET or of surfactants incorporated with MET sprays (S. Eskelsen, Adama US, personal communication). These are likely associated with environmental conditions at or immediately following applications, pre-treatment leaf cuticle development (Stover and Greene 2005), or concentrations effects of MET.

Effect of night temperature

Botton et al. (2011) suggests that fruitlet abscission is triggered by a critical level of carbohydrates within the fruit. If this is indeed the case, then the performance of MET as a fruit thinner will be dependent on several factors, including carbohydrate balance in the tree at the time of application. Solar radiation and temperature are the primary drivers of carbohydrate source, while sinks include structural wood, roots, leaves, fruit, and respiration. Competition between developing fruitlets in the corymb play an important role in abscission (Greene et al 2013), and when devoid of sufficient carbohydrate, likely initiate the abscission process. Temperature effects on dark respiration would therefore be a primary factor in determining the magnitude of this sink within developing fruitlets (Jackson 2003; Lakso et al. 2006). Indeed, Gonzalez et al. (2019, 2020) determined that the degree of fruit abscission with MET is highly dependent on night temperature, rate of application and cultivar, and suggests these are all explanations for the high variability that can exist in thinning efficacy among different trials. Gonzalez et al (2020) observed that when average night temperature exceeds 14 °C the day of and 4 d after a single application of MET, thinning efficacy is approximately 10%–25% higher in comparison with average night temperatures below 14 °C. This is corroborated by Byers (2002), who found low light conditions and periods of high night temperatures favour the abscission of fruitlets. Penzel and Kröling (2020) observed that a single application of MET at a relatively low rate (165 g ha⁻¹) effectively thinned RoHo 3615 apple trees under conditions of high night temperatures and low solar radiation. MET also appears more efficacious in warmer climates such as Israel (Stern 2014, 2015).

For comparison with the Gonzalez et al. (2020) study, key environmental conditions were summarized for the present study. Average nighttime air temperature between 2100 and 0700, cumulative solar radiation, and the percentage of time of hourly air temperatures above 14 °C were calculated for 0 to 4 d following each application of MET for each experiment using on-site research weather data (Table 18). In addition, the predicted thinning efficacy of MET based on proprietary web-based thinning software specifically designed for use with Brevis (15% MET; BreviSmart 2021; Adama Agricultural Solutions Ltd.), was included 1–4 d following application of MET (BreviSmart 2021). Briefly, the model predicts low, average, or high thinning efficacy based on cultivar, fruitlet size, solar radiation, and night temperatures based on studies done in Spain and MET trials from other European countries (Ton Besseling, Adama Agricultural Solutions Ltd., personal communication). The reduction in fruit per tree on trees treated with 330 to 350 mg L⁻¹ MET compared with the untreated trees (expressed as %) was calculated to provide an indication of thinning efficacy for each experiment. Overall, average nighttime temperature and accumulated solar radiation 4 d post treatment do not appear to be a good predictor of the thinning response to MET in this 4-yr study. When average nighttime air temperatures fell below 14 °C (30 May 2017, 31 May 2019), thinning on Gala and Honeycrisp was poor, but moderate on Ambrosia. The BreviSmart model predicted lower efficacy when applied on these days. On 31 May 2019, when nighttime temperature was on average 12.7 °C, predicted thinning efficacy was low, yet there was a 17% reduction in fruit on Ambrosia trees. When higher nighttime temperatures were experienced on 12 June 2017, MET was more effective in thinning Gala (22 mm fruitlet diameter) than Ambrosia (18.6 mm fruitlet diameter). Clearly, cultivars differ in their response to MET under similar environment conditions which cannot be explained by nighttime temperature and solar radiation levels and the BrevisSmart model. This study demonstrates stage of fruit development (fruitlet diameter) and concentration influence the thinning response to MET, but other factors including intra-fruit competition, tree carbohydrate balance, vis-à-vis differences in photoassimilate sources, and sinks, also influence the thinning response. Collectively, the interaction of these, and likely other unknown factors at this time, make it difficult to predict the overall thinning response to MET.

Table 18.

Year, cultivar, king fruitlet diameter, date of application of metamitron treatments, BreviSmart predicted thinning efficacy, average nighttime temperature, percent of time temperature was above 14°C, and solar radiation accumulated from the time of application to 4 d after application.