Factors Affecting Preharvest Fruit Drop of Apple Daniel Lee ......fruit in the canopy fell an...
Transcript of Factors Affecting Preharvest Fruit Drop of Apple Daniel Lee ......fruit in the canopy fell an...
Factors Affecting Preharvest Fruit Drop of Apple
Daniel Lee Ward
Virginia Polytechnic Institute and State University
Doctor of Philosophy
Horticulture
Richard P. Marini, Chair
John A. Barden
Eric P. Beers
Ross E. Byers
J. Roger Harris
John L. Hess
August, 2004
Blacksburg, Virginia
Keywords: abscission, cellulase, PGR, starch, seasonal, weather
Copyright 2004, Daniel Lee Ward
Factors Affecting Preharvest Fruit Drop of Apple
Daniel Lee Ward
(Abstract)
Apple preharvest fruit drop frequently results in severe economic losses. Cultural control of
preharvest drop has relied upon plant growth regulators (PGRs), but the loss of daminozide
(Alar) and 2,4,5-TP has severely limited the choices of effective stop-drop compounds. A more
complete understanding of factors involved in preharvest drop is therefore imperative.
Experiments were conducted to provide information about cellulase activity in the abscission
zone, effects of applied auxin and ethylene biosynthesis inhibition on drop, changing sensitivity
to abscission induction during the season, and relationships among seed number, fruit weight,
and day of drop. Observational studies were used to study effects of fruit maturity, canopy
positions, and morphology of stem attachment on time of fruit drop as well as characterizing the
natural timing of late-season fruit drop. Increased activity of cellulase, but not polygalacturonase,
in the abscission zone was detected within 4 days of cutting fruit to induce abscission. Both
aminoethoxyvinyl glycine (AVG) and naphthaleneacetic acid (NAA) applied 2 or 4 days after
cutting delayed drop, but NAA delayed drop 1.6 days longer than did AVG. Fruit of ‘RedChief
Delicious’(D) exhibited a significantly reduced sensitivity to abscission-inducing treatments from
mid-June until early July compared to earlier orr later in the season. Application of plant growth
regulators to cut fruit revealed a significant interaction of NAA treatment with AVG treatment
such that NAA delayed drop when applied with AVG but not without AVG. Fallen fruit had
lower starch and higher soluble solids than fruit on the tree on the day of collection. The highest
fruit in the canopy fell an average of 4.4d earlier than the lowest fruit. Day of drop was not
different for fruit from king blooms vs. side blooms within an inflorescence. There was a trend
for fruit from first year wood to drop later than fruit from older wood on ‘Delicious’, but not
‘Smoothee Golden Delicious’ trees. There was no detectable effect of angle of orientation of the
subtending spur on the limb, the pedicel:spur abscission zone, or fruit axis of symmetry on time
of fruit drop. No difference was detected in time of fruit drop between East and West or North
and South sides of the trees. No substantial variation in day of drop of individual fruit was
explained by number of seed in the fruit. Daily drop was recorded for three cultivars (‘RedChief
Delicious’, ‘Smoothee Golden Delicious’, and ‘Commander York’) for three years. Variance of
average day of drop from year to year was 40.1, while variance among cultivars within a year was
51.8. Variance from tree to tree within each cultivar, within each year, was only 18.6. Multiple
regression modeling to identify relationships between weather factors and daily fruit drop
revealed that much of the variability in time of drop was due to factors other than the weather
events modeled. The best regression models developed explained only 8% to 35% of the
variability in time of drop. The most important weather factors were daily minimum temperatures
and precipitation. Rain events of greater than 5.0 mm following a drier period appeared to cause
increased drop of all three cultivars in one out of the three years investigated.
iv
Acknowledgements
I want to thank my graduate committee for their support, patience and instruction throughout my
program. Many friends and family have been suppotive and inspirational to me during the years
of my PhD program. Thank you, may you be rewarded in kind.
v
Table of Contents
Abstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ii
Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iv
List of Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . vii
List of Figures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ix
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
Chapter One - Literature Review . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3Abscission: Definition, Scope, Model Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3Apple Fruit Abscission with Emphasis on Preharvest Fruit Drop . . . . . . . . . . . . . . . . . . . 8Horticultural Control Methods for Reducing Preharvest Drop . . . . . . . . . . . . . . . . . . . . 16
Chapter TwoCutting Apple Fruits Induces Cellulase Activity in the Fruit Abscission Zone . . . . . . . . 34
Chapter ThreeAbscission of cut Apple Fruits as Influenced by NAA and AVG Treatments and SeasonalChanges in Sensitivity to Abscission-inducing Treatments. . . . . . . . . . . . . . . . . . . . . . . 44
Chapter FourApple Fruit Ripening and Canopy Position Affects Time of Preharvest Drop. . . . . . . . . 54
Chapter FiveApple Fruit Attachment to the Tree Does Not Affect Time of Fruit Drop. . . . . . . . . . . . 66
Chapter SixEffect of Stigma Excision on Apple Fruit Set and Retention . . . . . . . . . . . . . . . . . . . . . 81
Chapter SevenRelationships Among Day of Year of Drop, Seed Number, and Weight of Mature AppleFruit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
Chapter EightTime Course and Effect of Environmental Factors on Preharvest Apple Fruit Drop. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107
General Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126
vi
Vita . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 132
vii
List of Tables
Table 1.1. Common names and chemical names of compounds effective at reducing preharvest
drop. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
Table 3.1. Ethylene concentration (mg@l-1) of headspace gas in jars containing three fruits
sampled 24 hours after treatment by cutting or spraying with 814 mg@l-1 ethephon. . . . . 50
Table 4.1. Mean starch index rating for fruit collected on the same date from the orchard floor or
from the tree by cultivar and date in 1995 and 1996 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
Table 4.2. Mean soluble solids concentration for fruit collected from the orchard floor or from
the tree by cultivar and date in 1996 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
Table 5.1. Percentage composition of the sample of fruits by cultivar [‘Smoothee Golden
Delicious' (GD), ‘RedChief Delicious’(D), and ‘Commander York' (Y)] and fruit per
spur, blossom type (king or lateral) or inflorescence type (terminal or lateral) . . . . . . . . 71
Table 5.2. Percentage composition of the sample of fruits by cultivar [‘Smoothee Golden
Delicious' (GD), ‘RedChief Delicious’(D)] and position of the spur, orientation of the
abscission zone (AZ) or orientation of the fruit in relation to the limb . . . . . . . . . . . . . . 72
Table 5.3. Mean day of year of drop of ‘Smoothee Golden Delicious' (GD), ‘RedChief
Delicious’(D), and ‘Commander York' (Y) fruit by number of fruit per spur, bloom type
(king or lateral), or inflorescence type (terminal or lateral) . . . . . . . . . . . . . . . . . . . . . . . 73
Table 5.4. Mean day of year of drop of ‘Smoothee Golden Delicious' (GD), and ‘RedChief
Delicious’(D) fruit by spur location, abscission zone (AZ) orientation, or fruit orientation
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74
Table 6.1. Mean number of filled seeds, unfilled seeds and wt. of mature dropped fruits from
viii
flowers with 0, 1, 2, 3, 4 or 5 stigmata excised before bloom in 1995 . . . . . . . . . . . . . . 87
Table 8.1. Mean (± standard deviation) day of the year of drop of apple fruit from six, seven, and
eight-year old trees of ‘Smoothee Golden Delicious'/M26 (GD), ‘RedChief
Delicious'/M26 (D), and ‘Commander York'/Mark (Y) . . . . . . . . . . . . . . . . . . . . . . . . 115
Table 8.2. Final multiple regression models and R2 for each year and cultivar . . . . . . . . . . . . . 116
ix
List of Figures
Figure 1.1. Relationship between percentage of ‘McIntosh' fruit dropped preharvest and tree yield11
Figure 1.2. Relationship between percentage of ‘McIntosh' fruit dropped preharvest and tree yield12
Fig. 2.1. Assay for CM-cellulase activity of fresh longitudinal sections through apple fruit
abscission zones sampled 0 (noncut), 2, 4, and 6 d after cutting fruits . . . . . . . . . . . . . . 40
Fig. 2.2. Frequency distribution plots and means of days from cutting to drop for ‘Delicious’
fruits on nontreated control limbs and on limbs sprayed with AVG (652 mg"L-1) or NAA
(10 mg"L-1) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
Fig. 3.1. Interaction plot of mean days from treatment to drop of cut fruits treated with NAA
(2000 mg@l-1) applied in lanolin to the seed cavity, with or without AVG (300 mg@l-1)
swabbed on the cut surface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
Fig. 3.2. Percentage of fruits abscised (out of 30 treated) 21 days after treatment by date of
treatment and treatment method. Treatments were: cutting (9, ª) or spraying with
ethephon (, •) to induce abscission . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
Fig. 4.1. Distribution of starch index ratings of fruit from two whole trees of ‘Delicious’ fruit
harvested 13 September 1996 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
Fig. 4.2. Scatterplot of day of drop(DOD) of highest and lowest fruit in canopy of ‘RedChief
Delicious’(RD), ‘Smoothee Golden Delicious’ (GD), and ‘Commander York’ (CY) in
1996 and 1997 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
Figure 5.1. Diagram of apple limb cross section with sampling locations for recording spur
orientation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
Figure 5.2. Diagram of apple limb cross section with sampling locations for recording
x
pedicel:spur abscission zone plane of orientation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
Figure 5.3. Diagram of apple limb cross section with sampling locations for recording fruit axis
orientation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
Figure 5.4. Scatterplot of day of drop (DOD) of fruit on the North and South sides of trees of
‘RedChief Delicious'(RD), ‘Smoothee Golden Delicious' (GD), and ‘Commander York'
(CY) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
Figure 5.5. Scatterplot of day of drop (DOD) of fruit on the East and West sides of trees of
‘RedChief Delicious'(RD), ‘Smoothee Golden Delicious' (GD), and ‘Commander York'
(CY) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
Fig. 6.1. Relationship between fruit set and number of stigmata excised before bloom in 1995
and 1997 by cultivar . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88
Fig. 6.2. Relationship between fruit retention and number of stigmata excised before bloom in
1995 and 1997 by cultivar . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
Fig. 6.3. Linear effect of number stigmata excised before bloom on early season fruit diameter
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
Fig. 7.1. The relationship between fruit weight and number of plump seeds per fruit for
‘Delicious’ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99
Fig. 7.2. The relationship between fruit weight and day of year of drop for ‘York’ . . . . . . . . 100
Fig. 7.3. Effect of number of fruit per tree on average day of year of drop of ‘Delicious’ fruit
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
Fig. 7.4. Relationship between fruit weight and number of plump seeds per fruit for ‘Delicious’
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102
xi
Fig. 7.5. Relationship between fruit weight and day of year of drop for three apple cultivars
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
Fig. 7.6. Frequency distributions for number of plump seeds per fruit . . . . . . . . . . . . . . . . . . . 105
Figure 8.1. Distribution of fruit drop as percentage of total crop across days for three years and
three cultivars . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118
Figure 8.2. R2 for percentage of total crop dropped by number of regressors included in the 10
best fit multiple regression models of each size for three years and three cultivars . . . 119
Figure 8.3. Mallows Cp for percentage of total crop dropped by number of regressors included in
the 10 best fit multiple regression models of each size for three years and three cultivars
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120
Figure 8.4. Smoothed surface plots of predicted percentage of total crop dropped by inches of
precipitation and average minimum temperatures over the seven days preceding drop for
three years and three cultivars . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121
Figure 8.5. Plots of predicted percentage of total crop dropped by inches of precipitation and
average minimum temperatures over the seven days preceding drop with rainfall greater
than 0.2 inches (dashed line) and without (solid line) for three years and three cultivars
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123
1
Introduction
Preharvest drop of apple is unpredictable and can result in severe economic losses in
some years. Some commercially important cultivars are seriously affected by this problem,
imposing a significant limitation on their profitability.
Preharvest drop has been a recognized problem in apple production for many years and
was discussed in early horticultural writings. Studies in the early part of the 20th century were
concerned with the physiology of fruit abscission as it occurred in the orchard. However, with the
advent of plant growth regulating chemicals (PGRs), the research emphasis changed dramatically
from physiology of abscission to the efficacy of PGRs. The discovery of the abscission delaying
properties of PGRs with auxin-like activity opened the way for development of chemical
horticultural tools to reduce preharvest drop losses. Since the use of stop-drop materials became
common practice, research on preharvest apple fruit abscission has centered on development of
better ways to use chemicals with commercial registrations. Unfortunately, chemicals as
horticultural tools have life spans limited by regulatory and economic variables outside the
control of apple growers. Research performed to refine the use of chemical stop-drops has done
little to increase our understanding of the underlying physiological processes.
Growers of cultivars with a propensity to drop would welcome other viable control
measures. In the future, breeders may be able to develop cultivars without preharvest drop
problems. Methods for creating such cultivars will require deeper understanding of the processes
involved in drop and of the differences among existing cultivars. In the near term, we may be
able to develop recommendations to reduce preharvest drop losses through more efficient use of
the chemical stop-drops or through other practices that can be implemented in existing orchards.
2
The goal of the investigations reported in this dissertation was to provide a rational basis
for a modification of current cultural practices. The experiments were designed to address a
range of questions surrounding the phenomenon of preharvest fruit drop. Some of the
experiments generated fundamental information (cellulase activity, auxin effect), and provide
useful knowledge and tools to increase our understanding of the physiological processes
involved. Results from other experiments, have potential short-term implications for decision
makers in the orchard (environmental effects, morphology of attachment, seasonality). All of the
experiments used apple trees in a research orchard to closely approximate the physical
environment to which we wish to make inferences.
Because of the relative paucity of information regarding the problem, the investigations
address a broad range of topics using a broad range of techniques. The cut-fruit model system
was used to provide information about cellulase activity in the abscission zone, effects of auxin
and ethylene biosynthesis inhibition on drop, and changing sensitivity to abscission induction
during the season. Interventions of stigma excision and shading were used to investigate
relationships among seed number, fruit weight, and day of drop. Observational studies were used
to study effects of fruit maturity, canopy positions, and morphology of stem attachment on time
of fruit drop as well as characterizing the natural timing of late-season fruit drop.
3
Chapter One - Literature Review
Abscission: Definition, Scope, Model Systems.
In the orchard apple fruit are lost to drop whether the cause of the drop is a blow from a
tractor or a physiological chain of events in the tree. Preharvest drop encompasses the fruit shed
by the tree and those lost due to acute external effects. Environmental physical processes like
buffeting winds as well as cultural operations such as mowing, spraying, summer pruning, and
harvesting when performed carelessly can all result in lost fruit. The physiological processes
associated with abscission will be the focus of this review with special emphasis on the
preharvest abscission of apple fruit and its control.
Abscission is the process of shedding organs by a plant. Sometimes this definition is
broadened to include other cell-separation processes (e.g. the separation of a yeast bud)
(Addicott, 1982). The former, more limited definition, is used in this review.
From the point of view of a plant two functions of abscission can be postulated:1)
facilitation of propagation, and 2) shedding an organ which has completed its useful performance
(Osborne, 1989). Abscission of mature apple fruit would facilitate propagation by increasing
dispersal of seeds by making them more easily available to frugivorous animals.
Addicott (1982) presents interesting discussion of evidence suggesting that abscission is a
fundamental process among plants extending back into the Devonian age. Processes that have
been longstanding in the phylogeny of plants would have more opportunities to participate in
advantageous mutations through the ages. Abscission has many roles in the adaptation of plants
to diverse ecological conditions. Therefore it is a response to many different stimuli and
presumably regulated in diverse ways.
4
Model plant systems in abscission experiments and cell separation processes in general
have been recently reviewed (Roberts, et al., 2000). Leaf abscission of bean explants (Phaseolus
vulgaris) has been long studied. Other plants that have been used for much laboratory research
are Coleus (Coleus blumei), impatiens (Impatiens walleriana), and elderberry (Sambucus nigra).
The model plant system Arabidopsis thaliana has become increasingly important in studies of
many types including abscission. Although Arabidopsis has small abscission zones and does not
exhibit fruit abscission the availability of etheylene response mutants and our extensive genomic
knowledge of Arabidopsis make it uniquely valuable as a model system.
Among fruit crops citrus, cherry, and more recently peach have been most studied by
abscission researchers. The motivation for these studies of fruit abscission has been the need to
stimulate abscission. Sweet and sour cherry fruit abscission has been more carefully
characterized than that of any other temperate zone fruit but, again the emphasis has been on
improving the efficiency of mechanical harvest (Bukovac, 1971).
Morphological and Anatomical Aspects of Abscission.
The separation events of abscission typically occur within a discrete, specialized tissue
known as the abscission zone (Addicott, 1982). The abscission zone (AZ) is a precisely
positioned zone of tissue that is specialized to allow for abscission. The plane of separation
typically passes through this abscission zone. The AZ is often a constricted region at the base of
the organ but may be indistinguishable from the surrounding tissue. Cell separation occurs in a
very restricted region within the AZ which is typically referred to as the separation layer. The so-
called separation layer may consist of several layers of cells.
A given organ may have several abscission zones associated with it. The leaves of
5
elderberry contain numerous abscission zones at the points of attachment of the leaflets to the
midrib (Osborne, 1989). The peduncle which subtends the peach fruit has three abscission zones
associated with it; at the base of the flower bud, between the peduncle and flower receptacle and
at the base of the fruit (Rascio, et al., 1985). The cherry fruit abscises at the juncture of the
pedicel and peduncle when immature, but when mature abscises at the juncture of the fruit and
receptacle (Wittenbach and Bukovac, 1972). The presence of independently regulated abscission
zones for the same organ suggests that the plant can use these different AZs in response to
different stimuli.
Generalizations about the anatomy of the AZ are difficult because the plant kingdom
contains widely divergent structures within abscission zones. Cells of the AZ have many juvenile
characteristics. They typically lack secondary wall thickenings and have little lignin or suberin.
The cells of the abscission zone are often densely cytoplasmic and smaller than surrounding cells
but most often they are not clearly distinguishable from surrounding cells. Ultrastructural
evidence has identified increases in dictyosomes and increased invaginated appearance of
plasmalemma of cells of the AZ (Sexton and Hall,1974). Tyloses form in vessels basipetal to the
AZ in some cases, presumably to reduce or prevent loss of sap during the healing of the wound
caused by abscission.
There is evidence of specialized AZ cells that must be present before abscission can
proceed (Osborne, 1976). AZ cells enlarge in response to ethylene but not auxin. This ethylene-
induced growth can be suppressed by applying auxin (Wright and Osborne, 1974). Cells with this
ability have been designated type II target cells, as distinct from typical type I parenchyma cells
which exhibit growth in response to applied auxin. This characteristic may not always be
6
included in the initial differentiation of a developing cell as mature cortical cells of bean leaf
explants can transdifferentiate into AZ responsive cells (McManus et al., 1998). Cell division is
not therefore a necessity for formation of competent AZ zones cells and perhaps also of
secondary AZs (adventitious abscission zones).
Biochemical Processes in Abscission.
Model systems for investigating abscission have elucidated many details of abscission.
The processes involved in cell separation including abscission have been recently reviewed
(Roberts, et al., 2002).
Degradation of cell wall structural components by hydrolytic enzymes is a critical part of
separation in the AZ. Increased ß-1,4-glucanase, or cellulase, synthesis has been detected in bean
explant AZs (Lewis and Varner, 1970). Other model plant systems exhibit similar increases in
cellulase associated with the catastrophic breakdown of cell walls during cell separation
(Roberts, et al., 2002). Increased polygalacturonase activity in the AZ has also been detected
(Bonghi, et al., 1992). Pectin methyl-esterase activity in the AZ also increases during abscission
(Sexton and Roberts, 1982). These enzymes and others likely work in conjunction to accomplish
abscission. The regulatory and degrading roles of these enzymes may be different among species,
but their characterization in model systems will hopefully provide the tools to study and
understand their roles in abscission in commercially important species.
Measurements of citrus fruit removal force after pulse stimulation with ethylene first
decreased then increased (Holm and Wilson, 1977). This “retightening” behavior indicates that
the separation processes involved are reversible. Likely a dynamic equilibrium of cellulose
synthases and cellulose hydrolases is maintained. This equilibrium could provide a point of
7
regulation of cell wall loosening.
The AZ eventually weakens from cell separation until the dead cells of the vascular
bundle are the only connected tissue. At this point a mechanical rupture of those vascular cells is
usually required. By increasing turgor pressure and loosening cell walls the cells of the AZ
expand during abscission which may provide the force to rupture the vascular strands. Often the
wind will combine with gravity to pull or shake off the fruit.
Regulation and Signaling in Abscission.
Many environmental stimuli can induce abscission. Leaf abscission is accelerated when
the plant is under stress (Addicott, 1982). Injury to an organ, whether by disease organisms,
sunburn, insect attack, or mechanical injury, is often followed by abscission. Hot weather can
also contribute to abscission (Wittwer, 1954). Programmed developmental events also initiate the
signal to abscise. The exact nature and perception of these signals and subsequent signal
transduction are poorly understood.
Hormonal regulation of abscission has been observed in many experimental systems
(Addicott, 1982). Much evidence has been reviewed describing the abscission-stimulating effects
of ethylene (Osborne, 1989, Abeles, 1992). Many experimental systems have relied on
exogenous ethylene to induce abscission (Sexton and Roberts, 1982). However, whether ethylene
is actually necessary for abscission has largely gone unexplored.
Ethylene and auxin are the plant hormones usually identified as most important in
regulating abscission. The abscission-inducing effect of ethylene and the inhibiting effect of
auxin provide two competing signals that can be manipulated (by experimenters or by the plant)
to control abscission. The mode(s) of action of ethylene are incompletely understood, but several
8
modes of action have been observed. Ethylene can induce increased synthesis of hydrolytic
enzymes in the AZ (Sexton and Roberts, 1982). Another mode of regulation of cellulase activity
in the AZ is the ethylene-stimulated secretion of cellulases (Abeles and Leather, 1971). Ethylene
inhibits the polar transport of auxin resulting in increased abscission (Beyer and Morgan, 1972).
Exogenous application of auxins can delay abscission induced by ethylene. Apparently an auxin
gradient must be maintained from the organ to the plant body to avert abscission (Addicott et al.,
1955). Diminished auxin production by an organ or decreased auxin transport may serve as a
signal to abscise.
Apple Fruit Abscission with Emphasis on Preharvest Fruit Drop
Apple fruit abscission and drop were topics studied using light microscopy and
histochemical techniques during the early and middle twentieth century. These anatomical and
biochemical results and observations are reviewed first, followed by a review of studies of
physiological, genetic, and horticultural factors associated with preharvest drop. Finally,
practices that have been used to reduce losses from preharvest drop are discussed.
Early-season fruit abscission (before June-drop, within 60 days after bloom) is a much
different process than preharvest drop. Abscission of the immature pedicels is preceded by cell
division in the abscission layer either at the zone of constriction or slightly distal to it (Heinicke,
1919, McCown, 1943). All tissues apparently undergo division during the process. The mature
pedicels (those surviving past June-drop) do not have cell division and differentiation of a
secondary abscission layer (McCown ,1943). Early and late season abscission are also regulated
differently as we know that sprays of NAA induce abscission of early season fruitlets but delay
abscission among late season fruit. Extrapolating from early season results to late-season
9
phenomena is likely of little value and potentially misleading so this review focuses on the late
season.
Description of the Abscission Zone of Apple Fruit
Apple fruit that drop before harvest typically have an attached pedicel and have abscised
from an AZ at the pedicel-spur junction. A constricted zone is usually visible between the pedicel
and cluster base of the spur with collenchyma extending through the area normally occupied by
fibers and stone cells (MacDaniel, 1936). The cortical region has little or no sclerenchyma and
the vascular cylinder has few fibers. The AZ has fewer xylem vessel elements and they have
scalariform pits instead of the normal round pits. The edges of the AZ are indefinite but the
region is usually about 0.5 mm of tissue between the cluster base and the pedicel. The separation
layer is not differentiated and cannot be distinguished from the rest of the AZ (in contrast to
many leaf AZs). The plane of fracture appears as a layer of living cells through the pith region
only 1 or 2 to 7 or 8 cells thick.
Preceding abscission the cells of the separation layer swell and become isotropic
(MacDaniels, 1936). Partial disintegration of the secondary walls of the separation layer likely
has begun by this stage. This is followed by dissolution of the pectic compounds of the middle
lamella. In cells of the other tissues of the AZ there is less disintegration of cellulose of the
secondary wall. The regulation of this disintegration is an important but poorly understood
component of apple fruit abscission.
Abscission is initiated in the pith and cortex independently. In ‘McIntosh' separation of
pith cells was initiated first whereas in ‘Golden Delicious', ‘Rome Beauty', ‘Fameuse', and
‘Delicious' pedicels cortical cell separation preceded separation in the pith (McCown, 1943).
10
This observation suggests that the variability in severity of drop problems among cultivars may
be due to structural and regulatory differences peculiar to individual cultivars.
After separation, the plane of fracture is smooth through the collenchyma and rougher
through the vascular tissue (MacDaniel, 1936). The roughness is likely the broken but
undegraded vessels of the AZ that were mechanically ruptured (McCown ,1943).The detachment
of the fruit leaves an open wound which the plant protects by forming cork after abscission. This
cork is formed by a cambium initiated a few cells proximal to the surface (McCown ,1943).
Physiological and Horticultural Factors Associated with Preharvest Apple Fruit Abscission
The severity of preharvest fruit drop varies greatly from year to year. Detailed review of
16 years records of preharvest fruit drop of ‘McIntosh’ revealed no significant correlation with
weather conditions (Southwick, 1938). The only weather effect that seemed related to drop was a
warm temperature shortly before harvest; apparently the effect was not substantial as the data
were not shown. These observations were based on drop records collected at irregular 2-4 day
intervals which may have obscured any weather effects.
Heavily fertilized or heavily mulched trees had a higher percentage of preharvest drop
than lightly fertilized trees (Southwick, 1938). For most of the trees evaluated, the percentage of
preharvest drop increased as the trees aged from 12 to 25 years. The aforementioned effects may
be attributable to the increased crop yield with heavy fertilizer, mulch, and age. The general
effect was a positive correlation between number of fruit borne by a tree and percentage of fruit
dropped preharvest (Fig. 1.1). A correlation between yield and percentage of fruit dropped was
also observed by Marsh, et al. (1959). This previously tabulated data are presented graphically to
emphasize this statistically significant relationship (Fig. 1.2).
11
y = 0.0674x + 0.8471
r2 = 0.5988
0
10
20
30
40
50
60
0 100 200 300 400 500 600
Yield (number of fruit / tree)
Per
cen
tag
e o
f fr
uit
dro
p p
reh
arve
st
Figure 1.1. Relationship between percentage of ‘McIntosh' fruit dropped preharvest and tree
yield. Data from Table 2 in Southwick (1938).
12
y = 1.0012x - 3.8941
r2 = 0.6524
0
10
20
30
40
50
60
0 10 20 30 40 50
Yield (boxes of fruit \ tree)
Per
cen
tag
e o
f fr
uit
dro
p p
reh
arve
st
Figure 1.2. Relationship between percentage of ‘McIntosh' fruit dropped preharvest and tree
yield. Data from Table 5 in Marsh et al. (1959).
13
High nitrogen nutrition status of the tree has been thought to be associated with greater
preharvest drop. ‘McIntosh’ trees that received nitrogenous fertilizer dropped a higher percentage
of their crop (the harvested crop was also larger for the fertilized trees) (Hoffman, 1940). When
fertilizer and moisture conditions were both high, the effect was more pronounced. Apparently
conditions leading to high nitrate availability near harvest time cause increased drop of
‘McIntosh’. Further investigation of this effect attempted to determine whether the nitrogen
effect, present at the level of the tree, was also present at the level of the spur. Spurs that held
their fruit longer were lower in nitrogen (Southwick,1940). Phosphorus followed the same trend.
Fruit that abscised early were borne on spurs with higher nitrogen content, but unfortunately all
spurs were sampled on the same day even though their fruits had abscised at different times. So
spurs whose fruit had abscised early had been without fruit for a longer time when sampled
which is confounded with the nitrogen effect. Whether nitrogen is directly responsible for
increased preharvest drop or whether this is an indirect effect of nitrogen increasing yield has not
be determined.
The seed content of fruit has been associated with the abscission of fruitlets early in the
season (Heinicke, 1917) and some work has extended this to the late-season drop (Southwick,
1938, and 1940, Marsh, et al., 1959). Though significant, the correlation between number of
seeds per fruit and drop r=0.231) explains only about 5% of the variation in time of drop of the
fruit (Southwick, 1938). Higher seed numbers could delay drop by a mechanism that occurs late
in the season; perhaps a hormonal signal is produced by the seeds. This seems unlikely because
late in the season the seeds are no longer connected to the vasculature of the plant and are living
free within the fruit. They could still be a source of a gaseous signal, but ethylene would likely
14
stimulate, not delay, abscission. It is more likely that the early season effects of higher seed
numbers result in developmental and structural differences that manifest themselves at the time
of preharvest drop.
A comparison of rootstocks for their effect on preharvest drop was made with ‘McIntosh’
and revealed no significant differences over a four-year period among standard, semi-dwarf and
dwarfing stocks (Southwick, 1938). This does not rule out the possibility that rootstocks that
cause heavier cropping may indirectly result in greater preharvest drop.
The influence of the root system on preharvest drop has been seen with root-pruning
experiments. Root pruning generally reduced preharvest drop on ‘Jonathan' trees, but the effect
was inconsistent from year to year (Ferree, 1992). In one year root pruning ‘Melrose’ trees after
June drop resulted in increased preharvest drop (Schupp and Ferree, 1987). While in other
experiments root pruning reduced abscission of ‘Melrose’ fruit (Schupp and Ferree, 1988).
Others have detected a significant reduction in preharvest drop following root pruning near time
of full bloom in two years with ‘McIntosh’, one of two years in ‘Cortland’, and in neither of two
years with ‘Delicious’ (Autio and Greene, 1999). The reduction in drop seemed more
pronounced for pruning treatments in mid June than in full bloom or dormant (Ferree, 1992).
Elving et al. (1991) reduced preharvest drop of ‘McIntosh’ with severe root pruning at full bloom
in a year with severe preharvest drop. The concomitant effects of root pruning in reducing
vegetative growth and yield may be the indirect cause of the reduced preharvest drop.
The question of why increased yield is associated with increased preharvest drop remains
unanswered. Likewise, why one fruit drops earlier than another is poorly explained by research to
date. Cultural practices could theoretically be modified in favor of reducing fruit prone to drop
15
early if we knew which fruit to target. If the one component of yield was clearly associated with
increased drop (for example fruit per spur) orchard managers could possibly adopt training and
pruning practices to reduced the proportion of fruit prone to drop.
The climacteric rise in ethylene that accompanies fruit ripening has been implicated in
drop. Exogenous application of ethylene as ethephon can induce abscission of apple fruit near the
time of harvest (Edgerton and Blanpied, 1970). As apple fruit ripen most cultivars exhibit a
climacteric with a dramatic increase in ethylene production. The notion that endogenous ethylene
is important in regulating fruit abscission seems reasonable and was investigated by observing
fruit internal ethylene concentration on the tree (Walsh, 1977). The climacteric increase in
internal ethylene occurred 4-10 days before drop in ‘Lodi' and 3-25 days before drop in
‘McIntosh'. Apparently the increase in ethylene does not uniformly affect fruit and likely other
regulatory mechanisms are responsible for much of the variability in time to drop after the
climacteric.
A recent discovery holds tremendous promise for apple production. Trees homozygous
for an allele of a ripening-specific 1-aminocyclopropane-1-carboxylate (ACC) synthase gene
produce less ethylene at the climacteric stage than wild-types (heterozygous trees) (Sato, et al.,
2004) . The homozygous clones also exhibit reduced preharvest drop. Breeders will now have the
possibility of screening new clones for reduced propensity to preharvest drop. Existing cultivars
will continue to need other means to reduce preharvest drop.
16
Horticultural Control Methods for Reducing Preharvest Drop
Before the advent of chemical plant growth regulators the losses from preharvest drop
could be reduced by harvesting early or planting cultivars with little tendency to drop fruit
preharvest. Unfortunately, suitable alternative cultivars are not always available and fruit
harvested too early are of lower market value. The development of cultural tools to limit or delay
preharvest drop was therefore of interest to growers and researchers. The literature on chemical
sprays for controlling preharvest drop have been reviewed by Edgerton (1973) and Wertheim
(1973).
In 1939 the first report of a plant growth regulating chemical effective for reducing
preharvest drop was published (Gardner, et al., 1939). This report of the effectiveness of plant
growth regulator sprays in reducing preharvest drop led to much activity in developing these
compounds as cultural tools. Many compounds have been investigated for this purpose (Marini et
al., 1989) and several have been found effective (Table 1.1). Three classes of compounds are
most important in the history of chemical preharvest drop control: 1) Auxins and compounds
with auxin-like activity, 2) SADH or daminozide , and 3) AVG, an ethylene biosynthesis
inhibitor. Each of these classes are reviewed in the chronological order in which they came into
use.
Auxins and Compounds with Auxin-like Activity
The auxins found to be most effective in reducing preharvest drop were the synthetic
NAA and NAAm. Auxins other than NAA (IBA, IAA, and indole propionic acid) were much
less effective than NAA, although they may have reduced drop a little compared to nontreated
controls (Gardner et al., 1940).
17
Table 1.1. Common names and chemical names of compounds effective at reducing preharvest
drop.
NAA naphthaleneacetic acid
NAAm naphthalene acetamide
Fenoprop, 2,4,5-TP 2-(2,4,5-trichlorophenoxy)propionic acid
Daminozide, SADH butanedioic acid mono(2,2-dimethylhydrazide)
Dicamba 3,6-dichloro-2-methoxybenzoic acid
Dichlorprop, 2,4-DP 2-(2,4-dichlorophenoxy) propanoic acid
AVG Aminoethoxyvinylglycine hydrochloride N-(phenylmethyl)-
1H-purine-6-amine
2,4-D (2,4-dichlorophenoxy) acetic acid
lactidichlor ethyl benzoic acid 3,6-dichloro-2 methoxy, 2-ethoxy,1-methyl, 2-
oxoethyl ester
triclopyr [(3,5,6-trichloro-2-pyridinyl)oxy] acetic acid
fenclopyr, CPPU N-(2-chloro-4-pyridyl)-N’-phenylurea
No common name triethanolamine salt of 2-(2,4,5-trichlorophenoxy)propionic
acid
No common name 2-methyl, 4-chlorophenoxyacetic acid
No common name, 2,4,5-TA triethanolamine salt of 2,4,5,-trichlorophenoxyacetic acid
No common name, 2,4,5-TAA 2,4,5-trichlorophenoxyacetamide
MCPB 4-(2-methyl-4-chlorophenoxy)butyric acid
18
Some compounds with auxin-like activity are also effective at reducing preharvest drop.
The common herbicide, 2,4-D is effective at reducing drop and has a more persistent effect than
NAA (Batjer and Thompson, 1946). Another phenoxyacetic acid, 2,4,5-TP, was found to be even
more effective as a stop-drop (Edgerton and Hoffman , 1951), but is somewhat slower (3-4 days)
in taking effect than NAA (Southwick et al., 1953). Dicamba has been determined to be at least
as effective as a stop-drop as other auxins (Marini and Byers, 1988), but its phytotoxic effects at
high concentrations (spur death and twig die back) make it an unattractive candidate for
registration (Marini et al., 1990). CPPU reduced drop for approximately eight days after
application, but by eleven days after application the effect had disappeared (Curry and Greene,
1993). A longer duration of effect was desired, so development of other compounds was pursued.
Sprays of NAA can control drop for one to two weeks after application (Murneek, 1954),
but when using 2,4,5-TP effective control can be achieved for 6 to 8 weeks (Batjer et al., 1954,
Murneek, 1954). Both 2,4,-D and 2,4,5-TP had a longer duration of effect and slower declination
of residue on the leaves than NAA (Looney and Cochrane, 1981). This longer lasting residue is
thought to be responsible for their more persistent activity. The longer duration of effect of 2,4-D
and 2,4,5-TP were valuable for use in late season cultivars, especially processing cultivars that
may be left on the tree longer than fresh market cultivars. These two compounds are no longer
registered for use in apples. An alternative means of obtaining more complete and long lasting
drop control is with repeated applications of NAA (Marini, et al., 1993).
Sprays have been the primary method of application of auxins used in commercial
orchards although other methods have been tried. Injections into the tree of 5 liters of 1000 ppm
NAA or NAAm had no effect on drop. Soil applied NAA or NAAm at 20 gallons per tree of 2.5,
19
5 or 10 ppm NAA were found to be only weakly effective (Gardner et al., 1940).
Experimental aerosol application was found to be highly effective (Tukey and Hamner,
1945) but is not practical for orchard use. NAA in the form of dust applied to take advantage of
the wetting by morning dew is equally effective as water-based sprays (Marth et al., 1945).
Airplane- delivered sprays have been used and are only slightly less effective than traditional
spraying (Thompson and Batjer, 1946). Because most growers can easily apply growth regulator
sprays in solution using ground equipment, this is the only method that has achieved widespread
use.
Along with timing of the spray, concentration of the auxin is a key factor in obtaining
satisfactory preharvest drop control. Concentrations in the range of 5 to 50 ppm are commonly
effective at reducing preharvest drop. With ‘Wealthy’ and ‘Delicious’ 10 ppm NAA was more
effective than five ppm in reducing preharvest drop (Murneek, 1940). In one year’s study with
four cultivars one ppm NAA proved too dilute to reduce drop (McCown and Burkholder, 1940).
A concentration-dependent effect of 2,4-D was seen on ‘Rome Beauty’, ‘Stayman’, and
‘Winesap’, but not on ‘Delicious’ (Marini, et al., 1988). Differential sensitivity of the cultivars
and the shorter time span recorded for drop among the ‘Delicious’ fruit could have been partly
responsible for this difference. Perhaps over a longer time, with increasing drop, differences
among concentrations would develop. 2,4,5-TP also has a documented concentration effect.
‘Delicious’ treated with 30 ppm 2,4,5-TP dropped less fruit than trees treated with 10 ppm
(Mattus and Moore, 1954).
Variability of response is a significant limitation of auxins as stop-drops. Variability of
response among cultivars has been noted. The effectiveness of the different compounds has been
20
seen to interact with cultivar by some (Looney, 1975), and they are not equally effective on all
cultivars. 2,4-D was effective on ‘Winesap’ and ‘Stayman Winesap’ but not on ‘Delicious’,
‘Oldenburg’, ‘York Imperial’, and ‘McIntosh’ (Batjer and Thompson, 1946, Harley, et al., 1946).
2-methyl,4-chlorophenoxyacetic acid was not effective on ‘McIntosh’(Edgerton and Hoffman ,
1951). Cultivars which initiate abscission in the cortex (‘Rome’ and ‘Delicious’) were more
responsive to sprays, than those which initiate abscission in the pith (‘McIntosh’ and ‘Jonathan’)
(Burkholder, and McCown. 1945). The mechanism of this selectivity is unknown; it may simply
be that the pith cells are deeper within the body of the plant and receive less exposure to the
applied compounds.
Auxins can also have undesirable side effects on the fruit. Early season cultivars (‘Close’,
‘Duchess of Oldenburg’, and ‘ Williams’) had hastened softening when treated with 10 ppm
NAA about 2 weeks before harvest (Batjer and Moon, 1945). ‘Summer Rambo’, ‘Jonathan’,
‘Delicious’, and ‘Rome Beauty’ exhibited a smaller ripening effect than the summer cultivars.
Softer fruit at harvest was seen in ‘McIntosh’, but only at a high NAA concentration (50 ppm)
when applied one month before harvest (Batjer et al., 1954). Methods developed to counteract
these effects are discussed below.
Daminozide
The first report of daminozide reducing preharvest drop was in 1966 (Edgerton and
Hoffman). Reduced drop was observed with applications from full bloom to a few days before
harvest, with the greatest response to application about one month before harvest (Edgerton,
1969). Even in a year with severe drop, daminozide provided preharvest drop control (Elving et
al., 1991).
21
Application of daminozide in consecutive years was not detrimental (Lord, 1971), but
caused an increase in return bloom the following year (Looney, et al. 1968).
Daminozide delays fruit softening and advances color development (Edgerton and
Hoffman, 1966, Elving, 1990). This was an important beneficial side effect of stop drop sprays of
daminozide. Unfortunately, daminozide is no longer registered in the United States.
Aminoethoxyvinylglycine, (AVG)
The modes of action of the various PGRs used for preharvest drop control are not
completely understood. AVG is known to be an ethylene biosynthesis inhibitor and this alone
makes it seem a likely tool for inhibiting abscission. AVG delays apple fruit drop as well or
better than auxin compounds (Bangerth, 1978). Nevertheless the specific action of AVG in
delaying drop is poorly understood. AVG applied to only the foliage caused a reduction in fruit
internal ethylene concentration, (Bangerth, 1978), indicating foliar uptake of AVG and
translocation of the AVG effect. Whether this is its mode of action (or its only mode of action)
for delaying drop is unknown.
AVG increased fruit removal force, (FRF), at harvest of ‘Golden Delicious' even though
the cultivar has minimal drop problems and no drops were recorded on treated or control trees
(Bangerth, 1978). It is possible that a different mechanism of loosening is responsible for this
difference as the lower strength of the AZ among control fruit was not accompanied by drop.
Applications of AVG have similar efficacy over a wide range of spray volumes (935 to
2805 l per ha) (Byers, 1997a). It is an easy compound to include in a spray program because it is
compatible in tank mixes with common pesticides (Byers, 1997b). Silicone-based surfactants
enhance the efficacy of AVG sprays (Byers, 1997a, 1997b), but other adjuvants (Regulaid,
22
Kinetic) do not.
Generally, the side effects of AVG stop-drop sprays on the fruit are not deleterious like
those of the compounds with auxin activity. Stop-drop sprays of AVG delay fruit softening
(Bangerth, 1978) and starch loss of fruit (Byers, 1997a, 1997b). This makes AVG more desirable
than the auxins which hasten fruit ripening. The delayed softening is an advantage especially
when harvest is being delayed. AVG also dramatically reduced watercore in ‘Delicious’ fruit
(Byers, 1997b).
Combinations and Interactions of PGRs for Preharvest Drop Control
Combinations of PGRs have been used commercially to: reduce preharvest drop, limit the
fruit softening effect of auxins, and increase fruit red color. The auxins, especially 2,4,5-TP,
cause fruit softening when applied as stop drops. These effects can be reduced by application of
maleic hydrazide (Smock, et al., 1951 and 1952). Making another PGR application to counteract
the undesirable side effect of auxins incurs an expense that would only be justified when
softening was a significant problem.
Under some conditions growers attempt to increase fruit red color by applying ethephon
to apple trees. When this was done fenoprop was the preferred auxin to inhibit fruit drop
(Looney, 1975). NAA can also be effective in controlling drop in ‘MacIntosh’ when ethephon is
applied to accelerate color development (Stover et al., 2003).
23
Literature Cited
Abeles, F.B. and G.R. Leather, 1971. Abscission: control of cellulase secretion by ethylene.
Planta. 97:87-91.
Abeles, F.B, P.W. Morgan, and M. E. Saltveit. 1992 Ethylene in Plant Biology.2nd ed. Academic
Press, New York.
Addicott, F.T. 1982. Abscission. Univ. Calif. Press., Berkeley, CA
Addicott, F.T., R.S. Lynch, and H.R. Carns. 1955. Auxin gradient theory of abscission
regulation. Science 121: 644-645.
Autio, W.R., and D.W. Greene. 1999. Effects of growth retarding treatments on apple tree
growth, fruit maturation and fruit abscission. J. Hort. Sci. 69:653-664.
Bangerth, F. 1978. The effect of a substituted amino acid on ethylene biosynthesis, respiration,
ripening and preharvest drop of apple fruits. J. Amer. Soc. Hort. Sci. 103:401-404.
Batjer, L.P. and H.H. Moon. 1945. Effect of napthaleneacetic acid spray on maturity of apples.
Proc. Amer. Soc. Hort. Sci. 46:113-117.
Batjer, L.P., H.W. Siegelman, B.L. Rogers, and F. Gerhardt. 1954. Results of four years’ tests of
24
the effect of 2,4,5-trichlorophenoxypropionic acid on maturity and fruit drop of apples in the
northwest. Proc. Amer. Soc. Hort. Sci. 64:215-221.
Batjer, L.P., and A.H. Thompson. 1946. Effects of 2,4-dichlororphenoxyacetic acid sprays in
controlling the harvest drop of several apple varieties. Proc. Amer. Soc. Hort. Sci. 47:35-38.
Beyer, E.M. Jr., and P.W. Morgan. 1972. Abscission: the role of ethylene in modification of
auxin transport. Plant Physiol. 48:202-212.
Bonghi, C, N. Rascio, A. Ramina, and G. Casadoro. 1992. Cellulase and polygalacturonase
involvement in the abscission of leaf and fruit explants of peach. Plant Mol. Biol. 20:839-848.
Bukovac, M.J. 1971. The nature and chemical promotion of abscission in maturing cherry fruit.
HortScience 6: 385-388.
Burkholder, C.L., and M. McCown. 1941. Effect of scoring and of A-napthyl acetic acid and
amide spray upon fruit set and of the spray upon pre-harvest fruit drop. Proc. Amer. Soc. Hort.
Sci. 38:117-120.
Byers, R.E. 1997a. Effects of Aminoethoxyvinylglycine (AVG) on preharvest fruit drop,
maturity, and cracking of several apple cultivars. J. Tree Fruit Prod. 2(1):77-97.
25
Byers, R.E. 1997b. Effects of Aminoethoxyvinylglycine (AVG) on preharvest fruit drop and
maturity of ‘Delicious' apples. J. Tree Fruit Prod. 2(1):53-76.
Curry, E.A., and D.W. Greene. 1993. CPPU influences fruit quality, fruit set, return bloom, and
preharvest drop of apples. HortScience. 28(2):115-119.
del Campillo, E., P.D. Reid, R. Sexton, and L.N. Lewis. 1990. Occurrence and localizing of 9.5
cellulase in abscising and nonabscissing tissues. Plant Cell 2: 245-254.
Edgerton, L.J. 1973. Control of abscission of apples with emphasis on thinning and pre-harvest
drop. Acta Hort. 34:333-343.
Edgerton, L.J., and G.D. Blanpied. 1970. Interaction of succinic acid 2,2-dimethyl hydrazide,
2-chloroethylphosphonic acid and auxins on maturity, quality and abscission of apples. J. Amer.
Soc. Hort. Sci. 95:664-666.
Edgerton, L.J. and W.J. Greenhalgh.1969. Regulation of growth, flowering and fruit abscission
with 2-chloroethanephosphonic acid. J. Amer. Soc. Hort. Sci. 94:11-13.
Edgerton, L.J., and M.B. Hoffman. 1951. The effectiveness of several growth regulating
chemicals in delaying the harvest drop of the McIntosh apple. Proc. Amer. Soc. Hort. Sci.
57:120-124.
26
Edgerton, L.J., and M.B. Hoffman. 1966. Inhibition of fruit drop and colour stimulation with
n-dimethylaminosuccinamic acid. Nature. 209:314-315.
Elving, D.C., E.C. Lougheed, C.L. Chu, and R.A. Cline.1990. Effects of daminozide,
paclobutrazol, and uniconazole treatments on ‘McIntosh' apples at harvest and following storage.
J. Amer. Soc. Hort. Sci. 115:750-756.
Elving, D.C., E.C. Lougheed, and R.A. Cline.1991. Daminozide, root pruning, trunk scoring, and
trunk ringing effects on fruit ripening and storage behavior of ‘McIntosh' apple. J. Amer. Soc.
Hort. Sci. 116:195-200.
Ferree, D.C. 1992. Time of root pruning influences vegetative growth, fruit size, biennial
bearing, and yield of ‘Jonathan' apple. J. Amer. Soc. Hort. Sci. 117:198-202.
Gardner, F.E., P.C. Marth, and L.P. Batjer. 1939. Spraying with plant growth substances to
prevent apple fruit dropping. Science. 90:208-209.
Gardner, F.E., P.C. Marth, and L.P. Batjer. 1940. Spraying with plant growth substances for
control of the pre-harvest drop of apples. Proc. Amer. Soc. Hort. Sci. 37:415-428.
Harley, C.P., H.H. Moon, L.O. Regeimbal, and E.L. Green. 1946. 2,4-dichlorophenoxyacetic acid
as a spray to reduce harvest fruit drop of apples. Proc. Amer. Soc. Hort. Sci. 47:39-43.
27
Heinicke, A.J. 1917. Factors influencing the abscission of flowers and partially developed fruits
of the apple (Pyrus malus L.). Cornell Univ. Agr. Expt. Sta. Bull. 393. pp. 41-114.
Heinicke, A.J. 1919. Concerning the shedding of flowers and fruits and other abscission
phenomena in apples and pears. Proc. Amer. Soc. Hort. Sci. 16:76-83.
Hoffman, M.B. 1940. The pre-harvest drop of mature McIntosh apples as influenced by
applications of nitrogen carrying fertilizers. Proc. Amer. Soc. Hort. Sci. 37:438-442.
HolmR.E, and W.C. Wilson. 1977. Ethylene and fruit loosening from combinations of citrus
abscission chemicals. J. Am. Soc. Hort. Sci. 102:576-579.
Lewis L.N., and J.E Varner. 1970. Synthesis of cellulase during abscission of Phaseolus vulgaris
leaf explants. Plant Physiol. 46:194-199.
Looney, N.E., D.V. Fisher, and J.E.W. Parsons. 1968. Some effects of annual applications of
n-dimethylaminosuccinamic acid (Alar) to apples. Proc. Amer. Soc. Hort. Sci. 91:18-24.
Looney, N.E. 1975. Control of ripening in ‘McIntosh' apples I. Some growth regulator effects on
preharvest drop and fruit quality at four harvest dates. J. Amer. Soc. Hort. Sci. 100:330-332.
Looney, N.E., and W.P. Cochrane. 1981. Relative effectiveness of, and residue declinations
28
values for dichlorprop, fenoprop and napthaleneacetic acid used to control preharvest drop of
McIntosh apples. Can. J. Plant Sci. 61:87-91.
Lord, W. 1971. Effects of annual sprays of succinic acid-2,2-dimethylhydrazide on vegetative
growth, fruiting, fruit quality and pre-harvest drop of ‘Delicious' apple trees. J. Amer. Soc. Hort.
Sci. 96(5):687-690.
MacDaniels, L.H. 1936. Some anatomical aspects of apple flower and fruit abscission. Proc.
Amer. Soc. Hort. Sci. 34:122-129.
Marini, R.P and R. Byers. 1988. Methods for evaluating chemical inhibitors of apple abscission.
HortScience. 23(5):849-851.
Marini, R.P., R.E. Byers, and D.L. Sowers. 1989. Growth regulators and herbicides for delaying
apple fruit abscission. HortScience. 24:957-959.
Marini, R.P., R.E. Byers, D.L. Sowers, and R.W. Young. 1990. Fruit abscission and fruit quality
of apples following use of dicamba to control preharvest drop. J. Amer. Soc. Hort. Sci.
115(3):390-394.
Marini, R.P., R.E. Byers, and D.L. Sowers. 1993. Repeated applications of NAA control
preharvest drop of 'Delicious' apples. J. Hort. Sci. 68:247-253.
29
Marsh Jr., H.V., F.W. Southwick, and W.D. Weeks. 1959. The influence of chemical thinners on
fruit set and size, seed development, and preharvest drop of apples. Proc. Amer. Soc. Hort. Sci.
75:5-21.
Marth, P.C., L.P. Batjer, and H.H. Moon. 1945. Relative effectiveness of sprays, dusts and
aerosols of naphthalene-acetic acid on harvest drop of apples. Proc. Amer. Soc. Hort. Sci.
46:109-112.
Mattus, G.E., and R.C. Moore. 1954. Preharvest growth regulator sprays on apples. I. Drop and
maturity 1952 and 1953. Proc . Amer. Soc. Hort. Sci. 64:199-208.
McCown, M. 1938. The mode of abscission of flowers and fruits of the apple. Ph.D. Diss. The
Ohio State Univ.
McCown , M. 1943. Anatomical and chemical aspects of abscission of fruits of the apple. Bot.
Gaz. 105:212-220.
McCown, M. and C.L. Burkholder. 1940. Very dilute a-naphthalene acetic acid spray and fruit
drop. Proc. Amer. Soc. Hort. Sci. 37: 429-431.
McManus, M.T., D.S. Thompson, C. Merriman, L. Lyne, and D.J. Osborne. 1998.
Transdifferentiation of mature cortical cells to functional abscission cells in bean. Plant Physiol.
30
116(3):891-899.
Murneek, A.E. 1940. Reduction and delay of fruit abscission by spraying with growth
substances. Proc. Amer. Soc. Hort. Sci. 37:432-434.
Murneek, A.E. 1954. 2,4,5-trichlorophenoxypropionic acid as a preharvest spray for apples. Proc.
Amer. Soc. Hort. Sci. 64:209-214.
Osborne, D.J. 1976. Auxin and ethylene and the control of cell growth. Identification of three
classes of target cells. In:Pilet, P. Ed. Plant growth regulation. Springer-Verlag. Berlin,
Germany.161-171.
Osborne, D.J. 1989. Abscission. Crit. Rev. Plant Sci. 8(2):103-129
Rascio, N., G. Casadoro, A. Ramina, and A. Masia. 1985. Structural and biochemical aspects of
peach fruit abscission (Prunus persica L. Batsch). Planta. 164:1-11.
Roberts J.A., K.A. Elliott, and Z.H. Gonzalez-Carranza. 2002 Abscission, dehiscence and
other cell separation processes. Ann. Rev. Plant Biol. 53:131-158
Roberts, J.A, C.A. Whitelaw, Z.H. Gonzalez-Carranza, and M.T. McManus. 2000. Cell
separation processes in plants - models, mechanisms and manipulation. Ann. Bot. 86:223-235.
31
Sato, T., T. Kudo, T. Akada, Y. Wakasa, M. Niizeki, and T. Harada. 2004. Allelotype of a
ripening-specific 1-aminocyclopropane-1-carboxylate synthase gene defines the rate of fruit drop
in apple. J. Amer. Soc. Hort. Sci.129:32-36.
Schupp, J.R., and D.C. Ferree. 1987. Effect of root pruning at different growth stages on growth
and fruiting of apple trees. HortScience. 22:387-390.
Schupp, J.R., and D.C. Ferree. 1988. Effects of root pruning at four levels of severity on growth
and yield of ‘Melrose’/M.26 apple trees. J. Amer. Soc. Hort. Sci. 113:194-198.
Sexton, R. and J.A. Roberts. 1982. Cell Biology of Abscission. Ann. Rev. Plant Physiol. 33:133-
162.
Sexton, R. and J.L. Hall. 1974 Fine structure and cytochemistry of the abscission zone cells of
Phaseolus leaves. 1. Ultrastructural changes occurring during abscission.Ann. Bot. 38(157):
849-854
Smock, R.M., L.J. Edgerton, and M.B. Hoffman. 1951. Some effects of maleic hydrazide on the
softening and respiration of apple fruits. Proc. Amer. Soc. Hort. Sci. 58:69-72.
Smock, R.M., L.J. Edgerton, and M.B. Hoffman. 1952. Inhibition of the ripening effect of certain
32
auxins on apples with maleic hydrazide. Proc. Amer. Soc. Hort. Sci. 60:184-192.
Southwick, L. 1938. Pre-harvest drop of the McIntosh apple. Master's Thesis. Mass. State
College, Amherst.
Southwick, L. 1940. Spur nitrogen and pre-harvest McIntosh drop. Proc. Amer. Soc. Hort. Sci.
37:435-437.
Southwick, F.W., I.E. Demoranville, and J.F. Anderson. 1953. The influence of some growth
regulating substances on pre-harvest drop, color, and maturity of apples. Proc. Amer. Soc. Hort.
Sci. 61:155-162.
Stover, E., M. J. Fargione, C. B. Watkins, and K. A. Iungerman. 2003. Harvest management of
Marshall 'McIntosh' apples: effects of AVG, NAA, ethephon, and summer pruning on preharvest
drop and fruit quality. HortScience. 38:1093-1099.
Thompson, A.H., and L.P. Batjer. 1946. The use of the airplane in applying hormone sprays for
the control of pre-harvest drop of apples. Proc. Amer. Soc. Hort. Sci. 47:44-48.
Tukey, H.B., and C.L. Hamner. 1945. Aerosol application of growth regulators to retard
abscission of apple fruits. Science. 101(2619):253-254.
33
Walsh, C.S. 1977. The relationship between endogenous ethylene and abscission of mature apple
fruits. J. Amer. Soc. Hort. Sci. 102:615-619.
Wertheim, S.J. 1973. Chemical control of flower and fruit abscission in apple and pear. Acta
Hort. 34:321-331.
Wittenbach, V.A. and M.J. Bukovac. 1972. An anatomical and histochemical study of abscission
in maturing sweet cherryfFruit. J. Amer. Soc. Hort. Sci. 97:214-219.
Wittwer, S. H. 1954. Control of flowering and fruit setting by plant regulators. In: Plant
Regulators in Agriculture. (H. B. Tukey ed.). Wiley, New York. pp. 62-80.
Wright, M. And D.J. Osborne. 1974. Abscission in Phaseolus vulgaris. The positional
differentiation and ethylene-induced expansion growth of specialized cells. Planta. 120:163.
34
Chapter Two
Cutting Apple Fruits Induces Cellulase Activity in the Fruit Abscission Zone
Abstract. Preharvest abscission of apple [Malus xdomestica (L.) Borkh.] fruits causes
significant crop loss in many years. In this study, fruit cutting was used to induce abscission in
August and September. Abscission zones of ‘Redchief Delicious’ Mercier strain fruits were
sampled 0, 2, 4, and 6 days after cutting. Thin-layer-plate assays were developed and used to
identify hydrolytic enzymes active in the abscission zone (AZ) after induction. Increased activity
of cellulase, but not polygalacturonase, was detected in the AZ following cutting. Cellulase
activity was consistently high in AZs 4 days after cutting. Both aminoethoxyvinyl glycine (AVG,
652 mg @L-1) and naphthaleneacetic acid (NAA, 10 mg @L-1) applied 2 or 4 days after cutting
delayed drop, but NAA delayed drop 1.6 days longer than did AVG. Fruits treated with AVG
dropped over a longer period than did control or NAA-treated fruits.
Apple fruit abscission shortly before harvest is responsible for substantial crop losses in
many years. Attempts to control this problem with plant growth regulators have yielded variable
results for unknown reasons. If the physiology of abscission was clearly understood, new
methods of controlling preharvest losses could be developed. All the physiological details of
abscission are not known for any system, but abscission of bean (Phaseolus vulgaris L.) leaf
petioles has been studied extensively. The degradation of cell well polysaccharides by
extracellular hydrolases is crucial during bean leaf abscission (del Campillo et al., 1990; Sexton
and Roberts, 1982) and, the degradation of cell walls in the abscission zone (AZ) is associated
35
with increased cellulase activity (Horton and Osborne, 1967). Ethylene treatments stimulate
increased synthesis and activity of cellulase and polygalacturonase (PGase) in the AZ of peach
fruit explants [Prunus persica (L.) Batsch.] (Bonghi et al., 1992). Pandita and Jindal (1991)
observed increased PGase and cellulase activity in the AZ of apple fruits induced to abscise with
ethephon (2-chloroethylphosphonic acid), an exogenous ethylene precursor.
Inducing fruit abscission by cutting apple fruits on the tree is a useful model system for
evaluating responses to growth regulators (Marini and Byers, 1988) and avoids both the
confounding effects of ethephon and the need to wait for natural drop. Whether this method of
abscission induction involves cellulase activity is unknown. The objectives of this study were;
1) to develop a simple means to evaluate enzyme activity in the AZ and use it to determine which
enzymes are active in cell wall hydrolysis during abscission of cut fruit; and 2) to determine if
growth regulators can delay abscission when applied after the appearance of hydrolytic activity in
the AZ.
Materials and Methods
Assays for hydrolytic enzyme activity. Trees used in this experiment were 8- and 9-year-
old ‘Redchief Delicious’ (Mercier strain) on M.26 rootstock. Each day from late July through
mid-September of 1996 and 1997 18-20 fruits were cut in half transversely through the seed
cavity. Five spurs were cut from the tree 0-6 days after fruit cutting to assay for enzyme activity,
and free-hand longitudinal sections through the abscission zone were cut with a disposable
microtome blade.
Assays, based on the work of Saleh-Rastin et al. (1991), were developed for visualizing
enzyme activity in the abscission zone. Thin layers (-2 mm thick) of 6% agarose in 50 mM
36
acetate buffer (pH 5.2) were prepared in 6-cm square petri dishes. To detect hydrolytic enzyme
activity, 0.2% substrate was included in the gel. The cellulase substrate was
carboxymethylcellulose and the PGase substrate was polygalacturonic acid. The cut surfaces of
spur pieces were pressed gently against the gel for 1 sec. Pieces of spur approximately 7 mm
long were pressed against the gel twice in each 1 cm square etched in the petri dish. The plates
were then sealed and incubated at 37 °C for 3 h. Activity of carboxymethyl-cellulase (CM-
cellulase) activity was visualized by staining the plates in a bath of 0.1% Congo red for 30 min
and destaining by rinsing several times with 1M NaCl. Areas of substrate hydrolysis were
detectable as unstained regions. Activity of PGase was visualized by staining for 20 min with
0.05% ruthenium red and destaining with several rinses of water.
Effects of applying growth regulators after induction. The experiment was conducted on
trees of 'Redchief Delicious’ (Mercier strain) on M.26 rootstock. All the fruits on 3 trees were
cut in half transversely through the seed cavity on 6 Sept. and sprays of AVG (652 mg @L-1 ) or
NAA (10 mg @L-1) based on acid equivalent, respectively, were applied to the point of drip with a
small hand sprayer 2 or 4 d after cutting. All sprays contained 0.2% Tween 20 (Sigma, St. Louis,
Mo.) as a surfactant. Non-treated controls (cut and noncut fruit) were included in the treatment
array. Abscised fruits were counted every day until all fruits had dropped and the number of days
from cutting to drop was calculated for each fruit.
The experimental design was a randomized complete block with three blocks. Whole
trees were blocks, 5 single scaffold limbs per tree were the experimental units and 10 fruits per
limb were subsamples. The treatment design was an augmented factorial array of two
compounds and two application dates augmented with a cut and an uncut control. The data were
37
analyzed using SAS’s MIXED Procedure with block as a random effect and treatment as a fixed
effect. Preplanned contrasts were used to test the hypotheses of interest.
Results and Discussion
Assays for hydrolytic enzyme activity. Because results were similar in both years, only
results from 1997 will be discussed. Cutting of fruits resulted in abscission within 7-12 d and
caused increased cellulase activity in the AZ (Fig. 2.1). PGase activity was not detectable in the
abscission zone (data not shown), but Pandita and Jindal (1991), using different assays, showed
that both cellulases and PGases were active in ethephon-induced abscission of apples. The
differences in season, environmental factors and method of induction could explain these
different results. Alternatively, transfer of PGase from the tissue to the agarose plate may have
been inadequate, or the levels of activity may have been below the level of detection for this
assay. We were able, however, to detect PGase activity in ripening apple fruits (data not shown).
Most of the cellulase activity was confined to a narrow band of cells presumably comprising the
separation layer (Fig. 2.1). The increased CM-cellulase activity of the AZ is consistent with
results obtained in bean (Horton and Osborne, 1967), peach (Ramina et al., 1993), and tomato
(Lycopersicon esculentum Mill.) (Tucker et al., 1993).
Cellulase activity increased following induction and was high in all AZ samples 4 d after
cutting (Fig. 2.1). On all sampling dates, the time course of cellulase activity development was
consistent. The 4-d delay between cutting and appearance of high levels of CM-cellulase activity
suggests the involvement of a multistage signal transduction pathway and de novo synthesis of
cellulase (Bonghi et al., 1992). Points of regulation along this pathway may provide appropriate
targets for manipulating apple fruit abscission.
38
The assays developed for this study were quick, simple, inexpensive and provide
information on tissue-level localization and relative activity of cellulase over time. The rapidity
of the assay allows the use of many samples directly from the orchard. These assays could be
used to identify enzyme activities of separating cells in other organs or at other times of the
season.
Effects of applying growth regulators after induction. The noncut control fruit dropped
more than 3 weeks later than did cut fruits (mean d to drop=36.7). The data were significantly
more variable than the other treatments. These data were therefore excluded from further analysis
and are not presented. Treatment with AVG or NAA delayed abscission of cut fruit (Fig. 2.2).
Data for treatment 2 and 4 d after cutting did not differ significantly, nor did date interact with
growth regulator treatment. Both compounds delayed fruit drop, but NAA delayed drop 1.6 d
longer than did AVG (Fig. 2.2).
The plate assay suggested that cellulase activity in the AZ increased 4 d after cutting
fruits. Growth regulator treatments were applied two days before and the day when increased
CM-cellulase activity was detected. Foliar sprays of both NAA and AVG delayed abscission
even after cellulase activity had increased (Fig. 2.2) but, NAA was more effective than AVG.
Commercial application of AVG to reduce preharvest drop is normally made 4 weeks before
anticipated harvest. In this study, treatments were applied near the time of harvest to investigate
the possibility of delaying abscission after the abscission process had begun. Applications of
NAA to reduce preharvest drop are often made as few as 7 to 10 d before anticipated harvest,
hence NAA may be more effective at delaying fruit drop late in the progress of developmentally
programmed abscission.
39
The distributions of the days to drop differed for all three treatments (Fig. 2.2). The
skewness coefficient is a non dimensional measure of the degree and direction of asymmetry of a
distribution. Treatment with AVG caused the distribution to be strongly skewed to the right
(skew=2.18), controls were skewed slightly to the right (skew=0.28), and the NAA treatment was
skewed slightly to the left (skew=-0.59) (Fig. 2). Many of the AVG-treated fruits fell shortly
after the controls dropped, with abscission of the remaining fruit spread over a longer time than
that of the NAA-treated fruits. Some fruits may have entered the climacteric and have already
begun to synthesize ethylene autocatalytically at the time of treatment. Application of AVG this
late may therefore have been less effective because of its reduced ability to attenuate ethylene
levels in fruits already synthesizing ethylene.
40
Fig. 2.1. Assay for CM-cellulase activity of fresh longitudinal sections through apple fruit
abscission zones sampled 0 (noncut), 2, 4, and 6 d after cutting fruits. Five samples are shown
per sampling date. Each sample was printed on a plate twice in the same square. The samples
were oriented with the pedicel toward the top of the gel. Fruits cut 15 August.
41
Treatment Mean
Control
. : : : : : . : : : : : : : . :--+---------+---------+---------+---------+---------+---
9.5 a
AVG
: : : : : : : : : : . . : : : : : : . : . : : : : : : : : : : . .--+---------+---------+---------+---------+---------+---
12.0 b
NAA
: : : : : : : : : : : : . . . : : : : : : : : : : : : : : : : .--+---------+---------+---------+---------+---------+--- 5 10 15 20 25 30
Days to drop
13.6 c
Fig. 2.2. Frequency distribution plots and means of days from cutting to drop for ‘Delicious’
fruits on nontreated control limbs and on limbs sprayed with AVG (652 mg"L-1) or NAA (10
mg"L-1). Means followed by different letters are significantly different at p#0.05 according to
contrasts. Each dot represents one fruit. Data for the 2 and 4 d treatments were pooled for AVG
and NAA treatment, as timing differences were nonsignificant.
42
Literature Cited
Bonghi, C., N. Rascio, A. Ramina, and G. Casadoro. 1992. Cellulase and polygalacturonase
involvement in the abscission of leaf and fruit explants of peach. Plant. Mol. Biol. 20:839-848.
del Campillo, E., P. D. Reid, R. Sexton, and L. N. Lewis. 1990. Occurrence and localization of
a 9.5 cellulase in abscising and nonabscissing tissues. Plant Cell 2:245-254.
Horton, R. F. and D. J. Osborne. 1967. Senescence, abscission and cellulase activity in
Phaseolus vulgaris. Nature 214:1086.
Marini, R. and R. Byers. 1988. Methods for evaluating chemical inhibitors of apple abscission.
HortScience 23:849-851.
Pandita, V. K. and K. K. Jindal. 1991. Enzymatic and anatomical changes in the abscission zone
cells of apple fruits induced by ethephon. Biol. Plant. 33:20-25.
Ramina, A., G. Casadoro, and N. Rascio. 1993. Structural, biochemical and molecular aspects of
abscission in peach. Acta Hort. 329:211-217.
Saleh-Rastin, N., M. Petersen, S. Coleman, and D. Hubbell. 1991. Rapid plate assay for
hydrolytic enzymes of Rhizobium. p.188. In: D. Keister and P. Cregan (eds.). The rhizosphere
43
and plant growth. Kluwer Academic. Dordrecht, The Netherlands.
Sexton, R. and J. A. Roberts. 1982. Cell biology of abscission. Annu. Rev. Plant Physiol.
33:133-162.
Tucker, M. G. Matters, S. Koehler, E. Kemmerer, and S. Baird. 1993. Hormonal and tissue-
specific regulation of cellulase gene expression in abscission. Curr. Plant Sci. Biotech. Agr.
16:265-271.
44
Chapter Three
Abscission of cut Apple Fruits as Influenced by NAA and AVG Treatments and Seasonal
Changes in Sensitivity to Abscission-inducing Treatments.
Abstract. The efficacy of ethephon as an apple thinner declines as fruits grow beyond 30 mm in
diameter, yet late in the season ethephon induces fruit abscission. Experiments were conducted
in 1996 and 1997 to identify the time of reduced sensitivity of ‘RedChief Delicious’ apple to
fruit abscission inducing treatments. Ethephon (2-chloroethylphosphonic acid) sprays at 814
mg@l-l or cutting fruit in half crosswise were used to induce fruit abscission on nine dates in 1996
and on six dates in 1997. Percentage of fruit abscised 21 days after treatment was calculated. On
three dates in 1997 fruit ethylene evolution 24-48 hours after treatment was measured. In both
years fruit exhibited a significantly reduced sensitivity to abscission-inducing treatments from
mid-June until early July. During the time of reduced sensitivity, cut fruits produced less
ethylene than ethephon-treated fruits. As sensitivity to abscission induction resumed, this
difference disappeared. Profiling changes in gene expression during the time of increasing
sensitivity to abscission induction may provide useful insight into the endogenous regulation of
preharvest fruit drop.
Manipulation of apple fruit abscission is important for reducing economic losses from
preharvest fruit drop. The loss of daminozide and 2,4,5-TP as stop-drop compounds has severely
limited the choices of chemicals available for preharvest drop control and emphasized the need
for more complete information about the preharvest abscission process. Abscission induction
45
methods used for investigating apple fruit abscission include spray applications of ethephon
(Pandita and Jindal, 1991) and the method of cutting fruits (Marini and Byers, 1988). Both
methods rely upon the consistent sensitivity of the fruits across time of treatment to provide
useful abscission responses.
The mode of action of cutting fruits to induce abscission is poorly understood. Non-
abscising leaves maintain an auxin gradient from leaf to stem (Addicott, et al., 1955). Supposing
a similar system controls abscission of apple fruit, cutting may collapse the auxin gradient by
removing the source of auxin or by disrupting the transport of auxin from the fruit.
Preliminary experiments indicated that cutting fruits in early summer failed to reliably
induce abscission. Therefore, further experiments were initiated to identify the time of the season
when ‘RedChief Delicious’ fruits have diminished sensitivity to abscission induction and to look
for differences in ethylene production by treated fruits during periods of sensitivity and
insensitivity to abscission induction.
The objectives of these experiments were: 1) to evaluate the abscission response to fruit
cutting or ethephon application throughout the growing season and, 2) to investigate the
hypothesis that an auxin gradient from the fruit to the spur prevents abscission by applying NAA
and/or AVG to the cut fruit surface.
Materials and Methods
Auxin and AVG Application to Cut Surface Experiment. Trees used in this experiment
were six and seven-year-old ‘Smoothee Golden Delicious'/M26 (GD), ‘RedChief Delicious’/M26
(D), and ‘Commander York'/Mark (Y) growing at the Virginia Tech College of Agriculture and
Life Sciences Kentland farm near Blacksburg, Va. Treatments were applied on 26 Aug. (GD and
46
D) and 28 Aug. (Y) in 1995, and on 11 Aug. ( all cvs) in 1996. Individual cut fruit were
experimental units and there were 10 replications per treatment, in a completely randomized
design, on each of the three cultivars. Four treatments were arranged in a 2 x2 factorial array:
1.)NAA in lanolin, 2.)NAA in lanolin plus AVG, 3.)lanolin with AVG, and 4.)lanolin. Emptied
seed cavities of cut fruits were filled with a suspension of lanolin containing NAA at the
concentration of 2g • l-1 with 0.2% Tween 20 to establish an auxin gradient from fruit to spur
(Mitchell and Livingston, 1968). A solution of 300 mg • l-1 AVG with 0.2% Tween 20 in water
was applied with a cotton swab to the entire cut surface to inhibit ethylene biosynthesis. All
fruits were covered with aluminum foil after treatment. Abscised fruits were counted daily until
all of the nontreated fruits had fallen from adjacent untreated trees. SAS’s GLM procedure was
used to perform analysis of variance (ANOVA) where days from treatment to drop was the
response variable for both years to test for effects of NAA, AVG and their interaction. The model
contained year, NAA and AVG as fixed effects. Significant interactions were further investigated
using the SLICE option to evaluate the effect of NAA within each level of AVG.
1996 Seasonality Experiment. Trees used in this experiment were 7-year-old ‘RedChief
Delicious’ Mercier strain/M.26, maintained using conventional commercial practices. Twenty
uniform limbs were selected and on each date each treatment was randomly assigned to three
limbs. On nine dates from 24 May through 4 Aug. 10 fruits on each of three limbs were labeled
and treatments were applied. The abscission inducing treatments were either cutting fruits in half
crosswise through the seeds or spraying the branch to the point of drip with ethephon at 814 mg@l-
1. Spray solution contained 0.2% v/v of the surfactant Induce (Helena Chemical Co., Memphis,
47
Tenn.). The experiment was a completely randomized design with three replications and ten
subsamples. The binomial response (abscised vs. not abscised) was recorded 21 DAT.
1997 Seasonality Experiment. Trees used in this experiment were 8-year-old ‘RedChief
Delicious’ Mercier strain/M.26, maintained using conventional commercial practices. On six
dates, chosen based on the results from 1996 (16 June - 30 July), five fruits on each of three
limbs were labeled and the limbs were treated. The experiment was a completely randomized
design with three replications and five subsamples. The treatments and data collection were the
same as used as in 1996.
Additional fruits treated on three dates (25 June, 11 July and 20 July) were sampled to
measure ethylene evolution. Three fruits from each limb were harvested 24 hours after
treatment. Each three-fruit sample was sealed in a 947 ml glass jar fitted with a rubber septum
and held at 22°C. After 6 or 24 hours, three one-ml subsamples were extracted with a syringe
from each jar. Ethylene was measured with a Shimadzu (GC-8A) gas chromatograph equipped
with an activated alumina column, flame ionization detector and a peak integrator. SAS’s GLM
procedure was used to perform an ANOVA for each date, and means were compared with
Tukey’s HSD.
Results and Discussion
Auxin and AVG Application to Cut Surface Experiment. There was a significant
interaction of NAA treatment with AVG treatment (P<0.0001) such that NAA delayed drop
when applied simultaneously with AVG (P<0.0001) but not without simultaneous application of
AVG (P=0.15, Fig. 3.1). NAA alone, delayed drop by 5 days. This is indirect evidence that apple
fruit, like other organs (Addicott, 1955), require an auxin gradient from the fruit to the point of
48
attachment to prevent abscission. Ethylene synthesis in response to cutting appears to induce
abscission even in the presence of an auxin gradient. The results from these experiments suggest
the hypothesis that endogenous IAA transport out of fruit is blocked by ethylene, resulting in
abscission. Preventing this ethylene effect is the probable mode of action of foliar AVG
applications used as a stop-drop tool.
Seasonality Experiments. In 1996 non-treated fruit did not abscise, so data for control
fruit are not presented. On 19 May and 8 June, and from mid-July through early August
abscission inducing treatments caused >30% of the fruits to abscise (Fig. 3.2). From mid-June
until early July, response to the abscission inducing treatments was attenuated and <30% of the
fruit abscised (Fig. 3.2). In 1997 treatments were applied during and after the period of reduced
sensitivity to abscission induction and abscission increased when fruit were treated after early
July (Fig. 3.2). The onset of sensitivity following reduced sensitivity occurred near the same
calendar date both years. This time is consistent with times when wounding caused fewer
‘Paulared’ and ‘Golden Delicious’ to abscise (Gucci, et al.,1991). In Massachusetts removing
the calyx end of ‘McIntosh’, but not ‘Golden Delicious’ fruits in early July hastened abscission
(Southwick, et al., 1962).
Both treatments induced abscission of nearly all fruits on dates when they were most
effective, but were less effective during June 1996. In 1997 abscission in response to treatments
was more pronounced, ranging from zero on 25 June and 2 July to 100% on 20 July (Fig. 3.2).
Profiling changes in gene expression during the transition from low to high sensitivity
(mid-June and early July) may provide a means to identify physiological changes leading to fruit
49
abscission.
During the period of reduced sensitivity to abscission induction (mid May to June) the
control and cut fruits released significantly less ethylene than the ethephon-treated fruits (Table
3.1). As sensitivity to abscission induction increased (11 July), the cut fruits began producing
more ethylene than the controls. The ability of the fruits to synthesize ethylene in response to
cutting appears to increase as they become more sensitive to abscission induction. On 20 July
ethylene evolution by cut fruits was similar to that of ethephon-treated fruits (Table 3.1).
We did not distinguish between the sources of ethylene released by the ethephon-treated
fruits. It could have been from the ethephon, the fruit itself or a combination. The source of
ethylene released by the ethephon-treated fruits on 25 June may have been entirely the ethephon
because wounding did not stimulate ethylene production. The higher levels of ethylene released
by the ethephon-treated fruit and cut fruit on 11 July and 20 July (Table 3.2) indicate that the
fruit were capable of synthesizing ethylene in response to the ethephon application and wounding
treatments on these dates. This would suggest a reduced ability to respond to ethylene, and not a
reduced ability to synthesize ethylene, during the period of reduced sensitivity to abscission
induction.
50
Table 3.1. Ethylene concentration (mg@l-1) of headspace gas in jars containing three fruits
sampled 24 hours after treatment by cutting or spraying with 814 mg@l-1 ethephon. Measurements
were made after jars were sealed for 24 hours for 25 June and after 6 hours for 11 and 20 Julyz,y.
Treatment date
Treatment 25 June 11 July 20 July
Ethephon 6.0 a 6.0 a 2.3 a
Cut 0.4 b 3.5 b 3.4 a
Control 0.3 b 0.0 c 0.4 b
z Each value is the mean of three replicates with three subsamples each.
y Means within a column followed by different letters are significantly different at " = 0.05,
according to Tukey’s HSD.
51
15
20
25
30
35
40
45
Day
s to
dro
p
- NAA
+ NAA
- AVG + AVG
Fig. 3.1. Interaction plot of mean days from treatment to drop of cut fruits treated with NAA
(2000 mg@l-1) applied in lanolin to the seed cavity, with or without AVG (300 mg@l-1) swabbed on
the cut surface. Each point represents the mean of 60 fruit.
52
0
20
40
60
80
100
19-May 8-Jun 28-Jun 18-Jul 7-Aug
Date of induction
Frui
t abs
ciss
ed a
fter
21
days
(%
)
Cut 1996
Ethephon 1996
Cut 1997
Ethephon 1997
Fig. 3.2. Percentage of fruits abscised (out of 30 treated) 21 days after treatment by date of treatment and treatment method.
Treatments were: cutting (9, ª) or spraying with ethephon (, •) to induce abscission. Full bloom occurred on 29 Apr. 1996 and 17
Apr. 1997.
53
Literature Cited
Addicott, F. T. R., R. S. Lynch, and H. R. Carns. 1955. Auxin gradient theory of abscission
regulation. Science 121:644-45.
Gucci, R., S. Mazzoleni, and F. G. Dennis Jr. 1991. Effect of fruit wounding and seed removal
on abscission of apple fruit between June drop and harvest. New Zeal. J. Crop Hort. Sci. 19:79-
85.
Marini, R. and R. Byers. 1988. Methods for evaluating chemical inhibitors of apple abscission.
Hortscience 23:849-851.
Mitchell, J.W., and G.A. Livingston. 1968. Methods of Studying Plant Hormones and Growth-
Regulating Substances. USDA Agr. Handbook No. 336.
Pandita, V. K. and K. K. Jindal. 1991. Enzymatic and anatomical changes in the abscission zone
cells of apple fruits induced by ethephon. Biol. Plant. 33:20-25.
Southwick, F. W., W. D. Weeks, E. Sawada, and J. F. Anderson. 1962. The influence of
chemical thinners and seeds on growth rate of apples. Proc. Amer. Soc. Hort. Sci. 80:33-42.
54
Chapter Four
Apple Fruit Ripening and Canopy Position Affects Time of Preharvest Drop.
Abstract. Improved understanding of factors controlling preharvest apple [Malus xdomestica (L.)
Borkh.] fruit abscission would enhance efficiency in using plant growth regulators and could
suggest new control methods. Experiments were conducted to determine whether fruits that drop
are riper than fruits on the tree at the same time, to determine if the highest and lowest fruits in
the canopy drop at different times and whether there were spatial patterns in fruit size and
maturity within the canopy. On several dates in two years, fruits from 7- and 8-year-old
‘RedChief Delicious’, (D) and ‘Commander York’, (Y) trees were collected from under the tree
or sampled from the tree. Starch index staining with potassium-iodide solution and soluble solids
concentrations were measured to estimate fruit maturity. In both years, with both cultivars, fallen
fruit had lower starch and higher soluble solids than fruit on the tree on the day of collection.
Another 2-year experiment used 7- and 8-year-old ‘Smoothee Golden Delicious’, D, and Y trees
to compare the day of drop of the 30 highest and 30 lowest fruit from each tree. In both years,
with all cultivars, the highest fruit in the canopy fell an average of 4.4d earlier than the lowest
fruit. There were no significant interactions of year or cultivar with tree height. Starch index
ratings and fruit size were determined for all fruit from two D trees sampled near harvest time.
No significant effects of canopy height, location on the scaffold limb (basal or distal half), or
their interaction on starch index were detected. Fruit from the basal half of limbs were
significantly larger than those from the distal half (156 g vs. 147 g).
55
Apple preharvest fruit drop results in severe economic losses in many years. Cultural
control of preharvest drop has relied upon plant growth regulators (PGRs), but the loss of
daminozide (Alar) and 2,4,5-TP has severely limited the choices of effective stop-drop
compounds. Poor understanding of factors controlling the abscission of mature fruit limits the
development of alternative control methods. Why one fruit and not another drops early is not
known. The climacteric rise in ethylene production by fruit on the tree is thought to trigger the
abscission process (Walsh, 1977). ‘Delicious’ fruit harvested weekly through the summer and
fall in NY and held in air containing 0.5 ml @ l-1 ethylene become competent to ripen (as
determined by starch staining) on 9 Oct. (Blanpied, 1972). Increased ethylene production by the
ripening fruit tissue and the establishment of an ethylene gradient between the fruit and the
pedicel coincides with the beginning of fruit drop (Blanpied, 1972).
Ripening fruit evolve ethylene which can induce abscission, but fruit that fall early may
be riper than those remaining on the tree or their pedicels may be weaker or more sensitive to the
ethylene signal from the fruit. If riper fruit abscise earlier, perhaps steps can be taken to
synchronize fruit maturity within the tree. Monitoring of fruit maturity may also provide a means
to predict preharvest drop severity.
The objectives of these experiments were: 1) to determine if fruit that fall early are riper
than those on the tree at the same time 2) to determine if there is a spatial pattern of fruit
maturation within the canopy, and 3) to determine if fruit from the top or bottom of the canopy
drop at different times.
Materials and Methods
On-tree vs. fallen fruit experiment. Trees used in this experiment were 7 and 8-yr-old
56
‘RedChief Delicious’ Mercier strain/M.26 (D) and ‘Commander York’/Mark (Y) maintained
using conventional commercial practices. In 1995 the orchard floor under two trees of each
cultivar was cleaned on 15 Sept. and 26 Sept. and fruit were allowed to drop for four days. On
19 and 30 Sept.(152 and 163 days after full bloom (DAFB) for D and 149 and 160 DAFB for Y)
the accumulated fruit were collected from under each tree and the same number of fruit that fell
was harvested from the tree. Sampling from the tree was done by randomly selecting a limb and
removing all the fruit along the limb until the appropriate number was obtained. Fruit were cut
in half, stained with an iodine-potassium iodide solution, and visually rated for starch content (1
indicating high starch and 8 indicating no starch) (Poapst, et al.,1959). In 1996 the orchard floor
was cleaned under four trees of each cultivar on 4 September and fruit were sampled at four-day
intervals (8, 12, 16, 20, and 24 Sept; 128- 148 DAFB for D and 125-145 DAFB for Y). Starch
index was recorded as in 1995. On 16 and 20 Sept. soluble solids concentration of juice
expressed from each sampled fruit was measured using a handheld refractometer (Atago, NC1).
Differences between mean starch index ratings and soluble solids concentration for each cultivar
at each date were analyzed using the TTEST procedure of the SAS System (ver. 7.0, SAS
Institute Inc., Cary, NC 27513).
Whole-tree fruit starch and fruit weight experiment. Two 8-year-old ‘RedChief
Delicious’ (Mercier strain)/M.26 trees maintained using conventional commercial practices were
used. On 13 Sept. 1996 (137 DAFB) all fruit were removed from the trees and each fruit was
weighed and indexed for starch content. Positions within the canopy were defined by numbering
the scaffolds acropetally and dividing each scaffold into basal and distal halves. The central
leader above the highest scaffold was regarded as the distal portion of a separate limb. Fruit from
57
each position were kept separate for weighing and starch rating. Fruit from limbs less than 2 cm
in diameter were pooled with the fruit from the basal half of the scaffold limb immediately basal
to them. Starch index ratings and fruit weights were subjected to a mixed effects analysis of
variance with tree as a random effect and limb and location as fixed effects. SAS’s MIXED
procedure was used to perform the analyses (Littel, et al., 1996).
Top vs. bottom of canopy fruit drop experiment. Trees used in this experiment were 7
and 8-year-old ‘Smoothee Golden Delicious'/M.26 (GD), ‘RedChief Delicious’ (Mercier
strain)/M.26, and ‘Commander York'/Mark maintained using conventional commercial practices.
In 1996 two trees per cultivar and in 1997 three trees per cultivar were used. Each year, in late
July, the 30 highest and 30 lowest fruit in the canopy of each tree were labeled with an indelible
marker. Fruit were collected from under each tree daily, and day of year of drop (DOD) was
recorded for each labeled fruit. SAS’s MIXED procedure was used to perform a mixed effects
model ANOVA with trees as a random effect and cultivar, year, and location in the tree (top or
bottom) as fixed effects.
Results
On-tree vs fallen fruit experiment. In both years the fallen fruit of both cultivars had
higher starch index ratings than fruit on the tree on the date of collection (Table 4.1). The fallen
fruit also had significantly higher soluble solids concentrations than those sampled from the tree
(Table 4.2).
Whole-tree fruit starch and fruit weight experiment. On 13 Sept., the mean starch index
rating of all the fruit on a tree was 3.8. The most frequently occurring starch rating was 3 but,
some fruit were observed with each of the eight ratings (Fig 4.1). The main effects of limb and
58
location and their interaction were nonsignificant (P>0.05) for the starch index rating (data not
shown). The mean fruit weight of all the fruit was 150 g. There was no significant effect of limb
or limb by location interaction for fruit weight but, locations were significantly different
(P=0.008). Fruit from the basal half of limbs were larger than those from the distal half (156 and
147 g, respectively).
Top vs bottom of canopy fruit drop experiment. Fruit from the top of the canopy fell 4.4 d
earlier than fruit from the bottom of the canopy (P=0.036) (Figure 4.2). There was no significant
interaction of location in the canopy with cultivar or year.
Discussion
Dropped fruit had higher starch ratings than fruit on the tree. Fruit higher in the canopy
fell earlier. However, canopy position did not influence mean starch rating.
Our observation that fruits that drop had less starch than fruit remaining on the tree agrees
with a previous report where ‘McIntosh’ fruit that dropped near harvest were low in starch
(Poapst, et al.,1959). The high soluble solids concentration in dropped fruit further supports the
notion that they were riper, not that they failed to accumulate starch. Preharvest drop appears to
be associated with ripening, as determined by low starch and high soluble solids. These results
suggest the hypotheses that ripening triggers preharvest abscission and it is the ripest fruit on the
tree that drop early.
Another indicator of ripening, the climacteric rise in ethylene production, was shown to
precede preharvest abscission (Walsh, 1997). A weak trend toward lower ethylene production
with increasing height in the canopy was noted in ‘Delicious’ (Farhoomand, et al., 1977). If
ethylene production triggers fruit drop, then one might expect that early drops would come from
59
low in the canopy, the opposite of what we observed.
Although only two trees were used, evaluation of starch of all the fruit on the tree failed
to detect any relationship between starch index ratings and tree height. This disagrees with a
previous report indicating that higher starch levels were associated with higher canopy levels in
‘Oregon Spur Delicious’ (Barritt, et al., 1987). Our method divided the canopy into smaller parts
and measured all the fruits on the tree so it is likely that our failure to detect any effect of canopy
height is because the trees we used were small and had relatively open canopies. Light levels may
have been more uniform throughout the canopy given the geographic location of the study. In the
eastern US canopy position effects on soluble solids (Campbell, 1991) and starch (Campbell and
Marini, 1992) are less pronounced than reported in the Pacific Northwest.
Apple fruit from the exterior and higher parts of the canopy were larger in size and weight
(Jackson, et al., 1971; Tustin, et al.1988; Looney et al., 1992). Our experiment failed to detect
any effect of canopy height on fruit weight, and the lower fruit weight of fruit from the distal half
of limbs is the opposite of the relationship seen in other studies.
‘Braeburn’ and ‘Granny Smith’ fruit borne laterally on one-year wood weigh less than
fruit borne on two-year wood (Volz, et al., 1993). Although we did not record age of wood upon
which fruit were borne, the lower weight of fruit on the distal half of limbs could have been
influenced by substantial one-year wood fruiting. The small size and open canopies of the trees in
our experiment may have less intra-canopy variation in light distribution which could reduce the
shading effects known to cause reduced fruit size (Jackson et al., 1971; Barritt, et al., 1987;
Tustin, et al., 1988).
60
Table 4.1. Mean starch index rating for fruit collected on the same date from the orchard floor or from the tree by cultivar and date in
1995 and 1996x,z,y.
Date of collection 1995 Date of collection 1996
Cultivar Location 19 Sept 30 Sept 8 Sept. 12 Sept. 16 Sept. 20 Sept. 24 Sept.
‘Delicious’ Tree 3.3 a 6.2 a 2.2 a 3.1 a 3.4 a 4.0 a 3.7 a
Floor 7.5 b 7.2 b 3.5 b 5.4 b 6.1 b 5.5 b 5.1 b
‘York’ Tree 1.2 a 2.3 a 1.1 a 1.3 a 1.4 a 1.3 a 2.7 a
Floor 2.7 b 6.3 b 2.3 b 2.9 b 4.2 b 5.0 b 5.3 b
z Each value is the mean of 10-111 observations
y Means within a cultivar and date followed by different letters are significantly different at " = 0.05 according to t-test.
x Starch rating of 1 is highest and 8 is lowest
61
Table 4.2. Mean soluble solids concentration for fruit collected from the orchard floor or from
the tree by cultivar and date in 1996z, y.
Date of collection
Cultivar Location 16 Sept. 20 Sept.
‘Delicious’ Tree 10.0 a 11.3 a
Floor 11.8 b 11.9 b
‘York’ Tree 9.8 a 9.6 a
Floor 11.3 b 10.9 b
z Each value is the mean of 10-40 observations.
y Means within a cultivar and date followed by different letters are significantly different at " =
0.05 according to t-test.
62
0
10
20
30
40
50 Pe
rcen
tage
of
tota
l
2 3 4 5 6 7 8Starch index rating
Fig. 4.1. Distribution of starch index ratings of fruit from two whole trees of ‘Delicious’ fruit
harvested 13 September 1996, n=986z.
Z Starch rating of 1 is highest and 8 is lowest
63
260
270
280
290
300
310 D
ay o
f dro
p -
high
est f
ruit
260 270 280 290 300 310 Day of drop - lowest fruit
CY96
GD96
RD96
CY97
GD97
RD97
Fig. 4.2. Scatterplot of day of drop(DOD) of highest and lowest fruit in canopy of ‘RedChief
Delicious’(RD), ‘Smoothee Golden Delicious’ (GD), and ‘Commander York’ (CY) in 1996 and
1997. Each point represents the mean DOD for fruit from a single tree (55-60 fruit). The
reference line of identity is included as a visual aid (if the DOD for the highest and lowest
samples from each tree were equal, the points would lie on the line).
64
Literature Cited
Barritt, B. H., C. R. Rom, K. R. Guelich, and M. A. Dilley. 1987. Canopy position and light
effects on spur, leaf, and fruit characteristics of ‘Delicious’ apple. HortScience 22:402-405.
Blanpied, G. D. 1972. A study of ethylene in apple, red raspberry, and cherry. Plant Physiol.
49:627-630.
Campbell, R. C. 1991. Canopy light environment influences apple leaf physiology and fruit
quality (photosynthesis). PhD. Diss. Virginia Polytechnic Institute and State University,
Blacksburg, VA.
Campbell, R.J. and R.P. Marini. 1992. Light environment and time of harvest affect ‘Delicious’
apple fruit quality characteristics. J. Amer. Soc. Hort. Sci. 117:551-557.
Farhoomand, M. B., M. E. Patterson, and C. L. Chu. 1977. The ripening pattern of ‘Delicious’
apples in relation to position on the tree. J. Amer. Soc. Hort. Sci. 102:771-774.
Jackson, J. E., R. O. Sharples, and J. W. Palmer. 1971. The influence of shade and within-tree
position on apple fruit size, colour and storage quality. J. Hort. Sci. 46:277-287.
Littel, R C., G.A. Milliken, W.W. Stroup, and R. Wolfinger. 1996. SAS System for Mixed
Models. SAS Publishing, Cary, NC. p.656.
65
Looney, N. E., R. L. Granger, C. L. Chu, L. N. Mander, and P. Pharis. 1992. Influences of
gibberellins A4, A4+7 and A4+iso-A7 on apple fruit quality and tree productivity. II. Other effects
on fruit quality and importance of fruit position within the tree canopy. J. Hort. Sci. 67:841-847.
Poapst, P.A., G.M. Ward, and W.R. Phillips. 1959. Maturation of McIntosh apples in relation to
starch loss and abscission. Can. J. Plant Sci. 39:257-263.
Tustin, D. S., P. M. Hirst, and I. J. Warrington. 1988. Influence of orientation and position of
fruiting laterals on canopy light penetration, yield, and fruit quality of ‘Granny Smith’ apple. J.
Amer. Soc. Hort. Sci. 113:693-699.
Volz, R. K., E. W. Hewett, and D. J. Wooley. 1993. Apple quality variation within the tree
canopy at harvest. Acta Hort. 343:56-58.
Walsh, C.S. 1977. The relationship between endogenous ethylene and abscission of mature apple
fruit. J. Amer. Soc. Hort. Sci. 102:615-619.
66
Chapter Five
Apple Fruit Attachment to the Tree Does Not Affect Time of Fruit Drop.
Abstract. Preharvest fruit drop of apple causes significant crop loss. Effects of fruit attachment
and location was assessed for their effect on day of year of drop of three cultivars [‘Smoothee
Golden Delicious'/M26 (GD), ‘RedChief Delicious’/M26 (D), and ‘Commander York'/Mark (Y)]
in one year. Day of fruit drop was recorded until all fruit had dropped. Day of drop was not
different for fruit from king blooms vs. lateral blooms within an inflorescence. There was a trend
for fruit on first year wood to drop later than fruit from older wood on D, but not GD trees. There
was no detectable effect of angle of orientation of the subtending spur on the limb, the
pedicel:spur abscission zone, or fruit axis of symmetry. No difference was detected in time of
drop between East and West or North and South sides of the trees. Variation in day of year of
drop among fruits of these three cultivars is likely more dependent on factors other than those
investigated in this study.
Shortly before harvest, apple fruit may drop to the orchard floor. This behavior varies
widely among cultivars and years but severe economic losses can result. Chemical stop-drop
sprays are often used to reduce the losses but, they are expensive, have regulatory risks (they can
quickly become unavailable), and are sometimes ineffective. Reducing drop through
modification of other cultural practices is a desirable alternative. If pruning or training practices
could be modified to reduce preharvest drop losses, they may provide a dependable cultural tool.
Research on apple preharvest fruit drop has largely ignored the effects of the morphology of how
67
individual fruit are attached to the tree. The short, stiff pedicels of some cultivars (e.g.
‘McIntosh’ and ‘York Imperial’) are thought to increase drop because fruit within a cluster touch,
so that as they grow one pushes off another. Whether other aspects of the fruit’s attachment
affect drop is unknown.
The objectives of this study were to determine whether the orientation of the spurs, the
orientation of the abscission zone (constricted area at the juncture of the pedicel and the cluster
base), or the orientation of the fruit affect time of fruit drop. We also wanted to determine if date
of fruit drop was affected by the number of fruit on a spur, position of the fruit within a cluster
(king vs. side fruit), and whether the inflorescence arose from one-year-old wood (lateral bud) or
older wood (terminal bud on spur). Data were also collected to determine whether fruit dropped
earlier on one side of the trees than another.
Materials and Methods
Trees used in this experiment were seven-year-old ‘Smoothee Golden Delicious'/M26
(GD), ‘RedChief Delicious’/M26 (D), and ‘Commander York'/Mark (Y) growing at the Virginia
Tech College of Agriculture and Life Sciences Kentland farm, near Blacksburg, Va.
Approximately one month before harvest, one limb was randomly selected on each of three trees
per cultivar. All the fruit on each limb were labeled and the day of year of drop for each fruit was
recorded by picking up the fallen fruit from under each tree daily until all the fruit had fallen. The
area under each tree was divided into quadrants to allow comparisons between different sides of
the tree.
For each fruit, we recorded whether it had been the central (king) blossom in the cluster
or a side blossom, whether it arose from a bud on one-year-old wood (lateral) or on older wood
68
(terminal), and the number of fruit on the spur. For the Y fruit, the type of blossom was not
readily identifiable.
The angle of attachment of the spur on which each fruit was borne was recorded to the
nearest 10 degrees (straight down from limb = 0º, straight up from limb = 180º) (Fig. 5.1).
Similarly, the angle of the plane of the pedicel:spur abscission zone (AZ) and the angle of the
axis of symmetry was recorded for each fruit (Figs. 5.2 and 5.3). The angle characteristics were
only recorded for the GD and D fruit because the short pedicels of the Y fruit made such
recording impractical. The angle data were reduced to categories of either top, bottom, or lateral
position of the spur, AZ, and fruit.
The composition of the sample was characterized using cross tabulations. Hypothesis
testing was performed to test for differences between the top, bottom, and lateral positions in day
of year of drop as well as the effects of fruit per spur, blossom type, inflorescence type and side
of tree. The MIXED procedure of the SAS System was used to perform analysis of variance with
tree and fruit in the model as random effects and position, type , or side as a fixed effect.
Results
The sampling method recorded the characteristics of all the fruit on the limb resulting in
different numbers of fruit for the different levels of each classifying variable. Most of the D fruit
were borne one to a spur with only 16% doubles and no spurs with three or 4 fruit (Table 5.1).
Most of the fruit were from king blossoms and from terminal buds. The GD and Y fruit were
mostly one or two fruit to a spur but, 18% of the GD and 6% of the Y were on spurs with three or
4 fruit. The lateral positions (accounting for 50% of the circumference of the limb) had the most
spurs with the top and bottom substantially lower (Table 5.2). The AZs and fruit angles were
69
distributed differently though with approximately two thirds of them being lateral and one third
located on top in relation to the limb. Only 3% and 8% of the AZs and fruit were located on the
top.
Day of year of drop was not different for fruit from king blooms vs. lateral blooms within
an inflorescence or from spurs with different numbers of fruit (Table 5.3). There was a notable
trend for fruit from first year wood to drop later than fruit from older wood on D, but not GD
trees (Table 5.3). There were too few lateral inflorescences in Y to make valid comparisons
(Table 5.1). There was no detectable effect of angle of orientation of the subtending spur on the
limb, the pedicel:spur abscission zone, or fruit axis of symmetry (Table 5.4). No difference was
detected in time of drop between East and West or North and South sides of the trees (Figs 5.4
and 5.5).
Discussion
Year-to-year drop patterns may vary for fruit attached in different ways therefore our
single year of results must be interpreted as preliminary. Variation in day of year of drop among
fruits of these three cultivars is likely more dependent on factors other than those investigated in
this study. Inherent limitations in the method of describing the attachment angles and position
may have lacked the necessary accuracy to uncover small or complex effects. The estimation of
the angles was simply done by eye and may have introduced appreciable error. Therefore, the
angles were categorized into top, lateral, and bottom, our intention being to simplify the
interpretation and to make it appropriate, given the inaccuracies in our angle measurements. We
are also missing orientation information in that the angle of attachment of the spurs, AZs, and
fruit only accounted for their position in a plane. Angles of AZs and fruit also may have changed
70
as the fruit continued to develop before drop.
‘Granny Smith’ fruit ripen earlier when borne on lateral inflorescences (Tustin, et al.,
1988). The earlier drop of fruit borne on lateral inflorescences may be related to earlier ripening
of those fruit.
Although the characteristics of attachment morphology evaluated here failed to uncover
any major determinants of time of drop, there are likely other morphological variations that
explain variation in time of drop. Side of tree may be a more important factor in older, larger
trees with more variability in canopy exposure. The challenge remains to determine whether the
earlier drop among heavily fruiting trees is reflected at the level of fruit per spur.
71
Table 5.1. Percentage composition of the sample of fruits by cultivar [‘Smoothee Golden
Delicious' (GD), ‘RedChief Delicious’(D), and ‘Commander York' (Y)] and fruit per spur,
blossom type (king or lateral) or inflorescence type (terminal or lateral).
Fruit per spur Blossom type Inflorescence typeCultivar 1 2 3 4 King Lateral Terminal Lateral
GD (N=399) 46 36 13 5 73 27 77 23D (N=330) 84 16 0 0 82 18 77 23Y (N=351) 50 44 5 1 - - 96 4
72
Table 5.2. Percentage composition of the sample of fruits by cultivar [‘Smoothee Golden Delicious' (GD), ‘RedChief Delicious’(D)]
and position of the spur, orientation of the abscission zone (AZ) or orientation of the fruit in relation to the limb. For GD N=399 and
for D N=330.
Spur location AZ orientation Fruit orientationCultivar Top Lateral Bottom Top Lateral Bottom Top Lateral Bottom
GD (N=399) 15 66 19 3 72 25 8 68 24D (N=330) 25 50 25 3 58 39 8 65 27
73
Table 5.3. Mean day of year of drop of ‘Smoothee Golden Delicious' (GD), ‘RedChief Delicious’(D), and ‘Commander York' (Y)
fruit by number of fruit per spur, bloom type (king or lateral), or inflorescence type (terminal or lateral). Standard error of mean in
parentheses.
Fruit per spur Blossom type Inflorescence typeCultivar 1 2 3 4 King Lateral Terminal Lateral
GD 301 (0.83) 302 (0.86) 301 (1.19) 302 (1.68) 301 (1.06) 301 (1.17) 301 (0.69) 301 (0.92)D 296 (3.49) 296 (3.98) - - 296 (3.16) 296 (3.57) 295 (3.05) 300 (3.36)Y 287 (1.10) 290 (1.14) 287 (3.59) 276 (6.47) - - 288 (0.79) 292 (3.76)
74
Table 5.4. Mean day of year of drop of ‘Smoothee Golden Delicious' (GD), and ‘RedChief Delicious’(D) fruit by spur location,
abscission zone (AZ) orientation, or fruit orientation. Standard error of mean in parentheses.
Spur location AZ orientation Fruit orientationCultivar Top Lateral Bottom Top Lateral Bottom Top Lateral Bottom
GD 301 (0.93) 301 (0.50) 301 (0.99) 301 (0.82) 301 (0.57) 296 (1.94) 302 (0.79) 301 (0.59) 299 (1.23)RD 296 (3.37) 295 (3.21) 296 (3.40) 295 (3.23) 295 (3.18) 294 (5.39) 295 (4.20) 296 (4.03) 295 (4.89)
75
Lateral
Bottom
Top
Lateral
Limb
Lateral
Bottom
Top
Lateral
Limb
Figure 5.1. Diagram of apple limb cross section with sampling locations for recording spur
orientation. For categorization spurs from the uppermost and lowermost 90 degree sections were
defined as top and bottom. All other spurs were pooled and called lateral.
76
Lateral
Bottom
Top
LateralLateral
Bottom
Top
Lateral
Figure 5.2. Diagram of apple limb cross section with sampling locations for recording
pedicel:spur abscission zone plane of orientation (dotted lines on the drawing). For
categorization planes approximately perpendicular to radii from the uppermost and lowermost 90
degree sections were defined as top and bottom. All other abscission zone planes were pooled
and called lateral.
77
Lateral
Bottom
Top
LateralLateral
Bottom
Top
Lateral
Figure 5.3. Diagram of apple limb cross section with sampling locations for recording fruit axis
orientation. For fruit location categorization, fruit whose axes of symmetry were in the
uppermost and lowermost 90 degree sections were defined as top and bottom. All other fruit
were pooled and called lateral.
78
285
290
295
300
305
285 290 295 300 305
Day of drop - North side
Day
of
dro
p -
So
uth
sid
e
GDRDCY
Figure 5.4. Scatterplot of day of drop (DOD) of fruit on the North and South sides of trees of
‘RedChief Delicious'(RD), ‘Smoothee Golden Delicious' (GD), and ‘Commander York' (CY).
Each point represents the mean DOD for all fruit from a single tree. The reference line of identity
is included as a visual aid (if the DOD for the North and South sides from each tree were equal,
the points would lie on the line).
79
285
290
295
300
305
285 290 295 300 305
Day of drop - East side
Day
of
dro
p -
Wes
t si
de
GDRDCY
Figure 5.5. Scatterplot of day of drop (DOD) of fruit on the East and West sides of trees of
‘RedChief Delicious'(RD), ‘Smoothee Golden Delicious' (GD), and ‘Commander York' (CY).
Each point represents the mean DOD for all fruit from a single tree. The reference line of identity
is included as a visual aid (if the DOD for the East and West sides from each tree were equal, the
points would lie on the line).
80
Literature Cited
Tustin, D. S., P. M. Hirst, and I. J. Warrington. 1988. Influence of orientation and position of
fruiting laterals on canopy light penetration, yield, and fruit quality of ‘Granny Smith’ apple. J.
Amer. Soc. Hort. Sci. 113(5):693-699.
81
Chapter Six
Effect of Stigma Excision on Apple Fruit Set and Retention
Abstract. Apple fruit with fewer seeds are less likely to set and be retained on the tree. An
experiment was carried out in two years on ‘Smoothee Golden Delicious'/M.26 (GD), ‘RedChief
Delicious’ Mercier strain/M.26 (D), and ‘Commander York'/Mark (Y). Flowers were manually
opened just before natural bloom and assigned to one of six treatments: 0,1,2,3,4,or 5 stigmata
excised using forceps. Fruit set and retention (beyond June-drop) were evaluated in both years.
Early season fruit size and mature fruit weight, number of filled seeds, and number of unfilled
seeds were recorded in only one year. Fruit size and mature weight, number of filled seeds and
number of unfilled seeds all decreased linearly with increasing number of excised stigmata.
Logistic regression revealed a significant quadratic effect of number of stigmata excised on set
and retention in both years that interacted with cultivar. GD fruit had a lower probability of set
and were less affected by treatment. D and Y fruit exhibited a sharp decline in probability of set
as number of excised stigmata increased from 4 to 5. Apparently successful pollination and
fertilization of even one stigma can result in significant fruit set and retention.
Following flowering many unpollinated and unfertilized flowers are shed by apple trees.
The importance of pollination to fruit set and retention is widely acknowledged.
‘Ben Davis’ fruit that drop during June-drop have fewer seeds and are smaller than fruit
from the same spur which persist (Sax, 1927). Yet at maturity, small and large ‘Ben Davis’ fruit
82
borne on the same spur are not different in seed content (Sax, 1927). The auxin content of seeds
varies throughout the growing season with times of seed high auxin content associated with
periods of very low fruit drop (Luckwill, 1953). Surgical removal of seeds causes fruit drop when
performed up to the time of June drop, but thereafter seed removal has little effect on drop
(Abbott, 1958). Removal of the some or all the stigmas of ‘McIntosh’ flowers before pollination
shows that the fruit can set with less than five functioning stigmata (Latimer, 1936).
Stigma excision before bloom was used to simulate the effect of incomplete pollination
and fertilization. The objectives of this study were to determine the effect of stigma excision on
fruit set, fruit retention, and early season fruit diameter.
Materials and Methods
Trees used in this experiment were 6 or 8 -year-old (1995 and 1997) Smoothee ‘Golden
Delicious'/M.26 (GD), RedChief ‘Delicious’ Mercier strain/M.26 (D), and Commander
‘York'/Mark (Y) maintained using conventional commercial practices. All experiments were
conducted at the Virginia Tech College of Agriculture and Life Sciences Kentland farm, near
Blacksburg, Va. Lateral flowers were removed from each spur and petals were removed from
unopened (“popcorn” stage) “king” flowers, which were tagged to receive one of six treatments
in a completely randomized design. Only flowers with five stigmata were used, and each flower
was considered an experimental unit. The six treatments were: pinching off 0, 1, 2, 3, 4, or 5
styles below the stigma from each of 25 flowers per cultivar. The remaining stigmata were hand-
pollinated on the 2 days following treatment. ‘Delicious’ flowers were pollinated with GD
pollen and Y and GD flowers were pollinated with D pollen. Treatments were applied on 13, 17,
83
and 19 Apr. 1995 to D, GD, and Y, respectively. Full bloom occurred on 16, 17 and 19 Apr. for
the D, GD, and Y trees, respectively.
To assess the effect of the treatments on fruit set (fruits that began growth) fruits attached
to the tree were counted 15, 18, and 21 days after treatment on D, GD and Y respectively.
Effects of the treatments on fruit retention (fruits remaining after “June drop”) were assessed by
counting attached fruits 48, 45 and 43 days after treatment on D, GD and Y respectively. In 1995
fruit diameter at the widest point was measured 69, 65 and 63 days after treatment on D, GD and
Y respectively. Mature fruit were collected when they dropped from the tree; then fruit weights
and the number of filled seeds and unfilled seeds were recorded.
Fruit set and retention data were analyzed using logistic regression to test for cultivar
effects and their interaction with linear and quadratic effects of number of excised stigmata. The
LOGISTIC procedure of the SAS System (ver. 8.2, SAS Institute Inc., Cary, NC 27513) was used
to perform the analysis with indicator variables for cultivars and the number of excised stigmata
as a regressor variable. Variables that contributed significantly to the fit (likelihood ratio P <
0.05), or that participated in a significant interaction, were retained in the model. To facilitate
interpretation of the logistic regression equations, predicted logits were calculated and converted
to probabilities for presentation. Early season fruit diameter, mature fruit weight, number of
filled seeds and number of unfilled seeds were analyzed using analysis of covariance to test for
linear trends across number of excised stigmata and homogeneity of slopes for the three cultivars.
Results
In both years the fruit set was decreased by stigma excision. There was a significant
84
quadratic by cultivar interaction effect on fruit set in both 1995 (P=0.0466) and in 1997
(P=0.0138)(Fig. 6.1). For all cultivars the probability of set declines steeply when number of
excised stigmata increases from 4 to 5. The set of GD fruit tended to be lower overall and less
affected by the excision treatments. In 1995 25% to 55% of fruit with all stigmata excised
remained on the tree until the time set was recorded, while in 1997 only 4% to 25% of such fruit
were remaining at time of fruit set.
Retention was similarly decreased by stigma excision in both years. There was a
significant quadratic by cultivar interaction effect on fruit retention in both 1995 (P=0.0194) and
in 1997 (P=0.0156)(Fig. 6.2). As with fruit set, fruit retention declined most dramatically as
number of excised stigmata increased from 4 to 5 for all cultivars. At the time of recording fruit
retention the differences among cultivars were less pronounced, but GD fruit were still less likely
to be retained than the others. In both 1995 and 1997 4% to 23% of fruit with all stigmata excised
were retained.
Analysis of covariance of early season fruit diameter identified a significant cultivar main
effect (P<0.0001) as well as a significant linear decrease with increasing number of stigmata
excised (P<0.0001), but the interaction effect explained no significant variation (P=0.3329). A
common slope model with separate intercepts for each cultivar was therefore fit (Fig. 6.3).
Fruit weight, number of filled seeds and number of unfilled seeds of mature apple fruit all
decreased linearly with increasing number of excised stigmata (Table 6.1).
Discussion
Our results with stigmata excision confirm that incomplete pollination and fertilization
85
result in decreased fruit set and retention as well as reduced early season fruit growth. The
excision of stigmata also caused reduced seed content of the fruit. The reduced seed complement
of the treated fruit may have indirectly caused the observed reduction in mature fruit size. Seeds
play a role in sink strength of growing fruit early in the season (Luckwill, 1953). Reducing the
seed content of fruit by excising stigmata may have caused those fruit to be weaker sinks for
photosynthate through much of the early growing season, thereby limiting their size potential.
Both fruit set and retention were little affected by removing 1-4 stigmata. This suggests
that incomplete pollination and fertilization may not always lead to early fruit abscission. Some
fruit were retained until maturity with only one functional stigma. ‘York’ fruit were especially
likely to remain on the tree even when stigmata had been excised. Competition among fruit on
the same spur would likely reduce the ability of incompletely fertilized fruit to remain on the tree
(Williams, 1977). Our method reduced each spur to exactly one fruit, so we eliminated the
within-spur competition among fruits that would usually be operative in a producing orchard.
A striking difference between the two years was the change from fruit set to fruit
retention. In 1997, but not 1995, very few fruit fell between the times when we collected fruit set
and fruit retention data. Our method of recording fruit set and retention had a limitation, in that
we may have evaluated set at a developmentally more advanced stage in the second year.
Additionally, the method of pollination was different between the two years. The timing of the
abscission of unpollinated flowers is variable from year to year and may have been part of the
cause of the differences in set between the two years. There were overall differences in set
between the two years; however, the proportion of fruit that were retained when all their stigmata
86
had been excised was very similar in the two years.
An unusual finding was that even though some fruit had all their stigmata removed before
bloom they still had seeds at maturity. Similar findings been attributed to germination of pollen
on the stumps of the style remaining after removing the stigma (Namikawa, 1923).
Stigmata excision as a model for incomplete pollination and fertilization has limitations,
but provides an easily manipulable system. It has permitted us to establish a range of seed
numbers without causing the fruit with small numbers of seeds to be abscised before maturity.
The results also corroborate the finding that fruit can grow to maturity with few seeds when intra-
spur competition is not strong (Williams, 1977) .
87
Table 6.1. Mean number of filled seeds, unfilled seeds and wt. of mature dropped fruits from flowers with 0, 1, 2, 3, 4 or 5 stigmata
excised before bloom in 1995z. Linear trends are based on individual fruits for each response variable with number of excised
stigmata as the regressor variable.
‘Delicious’ ‘Golden Delicious’ ‘York’
Stigmata
excised
n Filled
seeds
Unfilled
seeds
Fruit wt
(g)
n Filled
seeds
Unfilled
seeds
Fruit wt
(g)
n Filled
seeds
Unfilled
seeds
Fruit wt
(g)
0 12 8.8 0.7 221 18 8.2 1.4 210 16 8.2 0.7 126
1 11 7.8 1.0 200 20 7.7 1.8 211 16 7.8 1.3 108
2 9 5.8 2.2 198 15 7.7 2.3 200 18 7.6 1.6 123
3 7 5.4 1.6 204 13 6.8 2.5 188 16 6.7 2.8 126
4 11 2.8 1.9 175 8 5.0 4.0 206 18 6.4 1.9 104
5 0 - - - 4 2.8 4.0 186 0 - - -
Linear trend
Pr > F .0001 .019 .0013 .0001 .0001 .14 .0027 .0007 .0217
r2 .75 .11 .19 .32 .20 .028 .11 .14 .19
z Each value is the mean of the number of observations indicated by n (out of 25 treated blossoms).
88
0.0
0.2
0.4
0.6
0.8
1.0
0 1 2 3 4 5 6
1995
Pro
bab
ility
of
fru
it s
et
0.0
0.2
0.4
0.6
0.8
1.0
0 1 2 3 4 5 6
Stigmata excised
YRDGDY obsRD obsGD obs
1997
Fig. 6.1. Relationship between fruit set and number of stigmata excised before bloom in 1995
and 1997 by cultivar. Predicted quadratic curves and observed proportions are shown for
‘RedChief Delicious’ (), ‘Smoothee Golden Delicious’ (±), and ‘Commander York’ (•)
apples. Each data point is the proportion of fruit set out of 25 trials.
89
0.0
0.2
0.4
0.6
0.8
1.0
0 1 2 3 4 5 6
1995
0.0
0.2
0.4
0.6
0.8
1.0
0 1 2 3 4 5 6Stigmata excised
YRDGDY obsRD obsGD obs
1997
Pro
bab
ility
of
fru
it r
eten
tio
n
Fig. 6.2. Relationship between fruit retention and number of stigmata excised before bloom in
1995 and 1997 by cultivar. Predicted quadratic curves and observed proportions are shown for
‘RedChief Delicious’ (), ‘Smoothee Golden Delicious’ (±), and ‘Commander York’ (•)
apples. Each data point is the proportion of fruit set out of 25 trials.
90
35
40
45
50 F
ruit
diam
eter
(m
m)
Stigmata excised1 2 3 4 50
Fig. 6.3. Linear effect of number stigmata excised before bloom on early season fruit diameter
for ‘RedChief Delicious’ (), ‘Smoothee Golden Delicious’ (±), and ‘Commander York’ (•)
apple fruit in 1995.
91
Literature Cited
Abbott, D.L. 1958. The effects of seed removal on the growth of apple fruitlets. Annual Report,
Long Ashton Research Station. pp. 52-56.
Latimer, L.P. 1936. The effect of reducing the number of functioning stigmas on fruit-setting and
characteristics of the McIntosh apple. Proc. Amer. Soc. Hort. Sci. 34:22-25.
Luckwill, L.C. 1953. Studies of fruit development in relation to plant hormones. I: Hormone
production by the developing apple seed in relation to fruit drop. J. Hort. Sci. 28:14-24.
Namikawa, I. 1923. Growth of pollen tubes in self-pollinated apple flowers. Bot. Gaz. 76:302-
310.
Sax, K. 1927, Factors affecting fruit development of the apple. Me. Agr. Exp. Sat. Bull. No. 298.
Studies in orchard management.pp. 53-84.
Williams, M.W. 1977. Adverse weather and fruit thinning chemicals can affect seed content and
size of ‘Red Delicious’ apples - What can growers do about it? Proc. Washington State Hort.
Assoc. 73:157-161.
92
Chapter Seven
Relationships Among Day of Year of Drop, Seed Number, and Weight of Mature Apple Fruit
Abstract. Preharvest fruit drop of apple [Malus xdomestica (L.) Borkh.] can cause significant
crop losses, but factors controlling date of individual fruit drop are unknown. In three types of
experiments, we investigated the relationships among seed number per fruit, fruit weight, and day
of year of drop. By shading in mid-May and stigma excision before bloom we induced
variability in seed number. Dropped fruit were were collected daily from late August until all
fruit had dropped and weighed, and their seeds were counted. Regression analyses were used to
assess relationships among day of drop, fruit weight, and seed number per fruit. Substantial
variation in day of drop of individual fruit was not explained by number of seeds per fruit in
these experiments with ‘Smoothee Golden Delicious’, ‘RedChief Delicious’, and ‘Commander
York’.
Preharvest fruit drop often causes severe crop losses of several important apple cultivars
of the northeast and mid-Atlantic states. The problem has been controlled partially using plant
growth regulators (PGRs). Factors affecting response to PGRs and seasonal variation are poorly
understood. Even less is known about factors determining which of the fruit on a tree abscise
first. Although Walsh (1977) concluded that the fruit of ‘Lodi’ and ‘McIntosh’ drop shortly after
the onset of the climacteric, the time from climacteric rise to drop varied from 3 to 25 d. In the
post-bloom period, fruit set and retention is related to hormone production by the seeds (Dennis,
93
1967; Luckwill, 1953). Seed number per fruit has been implicated in the developmental
regulation of apple fruit growth and final size (Murneek and Schowengerhdt, 1935; Williams,
1977). Higher seed numbers also were correlated with later drop in mature ‘McIntosh’ apples
(Southwick, 1938a, b), but seed number explained only a small amount of the variation in time of
drop. Whether or not seed number affects the time of preharvest drop for other cultivars is
unknown.
If seed number does affect the time of drop, preharvest drop losses could be predicted by
sampling fruit to determine seed number, and thinning practices that affect seed number could be
modified to reduce losses. Similarly, any effect of seed number on fruit weight may be important
to growers choosing among available thinning options. A relationship between fruit weight and
date of drop may be useful for evaluating the cost of preharvest losses and deciding upon
appropriate control measures.
The objectives of these experiments were to investigate the relationships among seed
number, day of year of drop, and weight of individual apple fruit of three commercially important
cultivars having varying tendencies for preharvest drop problems.
Materials and Methods
Stigma-excision experiment. Trees used in this experiment were 6-year-old ‘Smoothee
Golden Delicious'/M.26 (GD), ‘RedChief Delicious’ Mercier strain/M.26 (D), and ‘Commander
York'/Mark (Y) maintained using conventional commercial practices. All experiments were
conducted at the Virginia Tech College of Agriculture and Life Sciences Kentland farm, near
Blacksburg, Va. Full bloom occurred on 16, 17 and 19 Apr. for the D, GD, and Y trees,
94
respectively (days 106, 107, and 109 of the year, respectively). Lateral flowers were removed
from each spur and petals were removed from unopened (“popcorn” stage) “king” flowers, which
were tagged to receive one of six treatments in a completely randomized design. Only flowers
with five stigmata were used, and each flower was considered an experimental unit. The six
treatments were: pinching off 0, 1, 2, 3, 4, or 5 styles below the stigma from each of 25 flowers
per cultivar. The remaining stigmata were hand-pollinated on the 2 days following treatment.
‘Delicious’ flowers were pollinated with GD pollen and Y and GD flowers were pollinated with
D pollen. Treatments were applied on 13, 17, and 19 Apr. 1995 to D, GD, and Y, respectively.
Beginning 26 Aug., fruit were collected daily as they fell from the tree until all fruit had fallen.
Fruit collected on a given day were placed in a paper bag and stored in conventional cold storage
at 2 °C until evaluation. The weight of each fruit was recorded. Fruit were cut open and the
number of plump seeds and aborted seeds were counted. Fruit without an attached pedicel were
considered to have been pushed off by other fruits and were omitted from the analysis.
Data from all treatments were pooled, and regression analyses were conducted using the
REG procedure of the SAS system (SAS Institute Inc, Cary, N.C.). The regressors (plump seed
number, aborted seed number, and individual fruit weight) were evaluated for their effects on
time of drop. Plump seed number and aborted seed number were evaluated for their effects on
individual fruit weight.
Shading experiment. The experimental units were 6-year-old ‘RedChief Delicious’
Campbell strain/M.26 trees maintained using conventional commercial practices. Trees were
blocked on uniformity of bloom, and one tree in each of two blocks was randomly assigned to
95
receive one of four treatments. Trees were completely covered with 92% black plastic shade
fabric (saran) for 0, 1, 2, or 3 d. Shade was imposed beginning 12 May when mean fruit
diameter was about 13 mm. The experimental design was a randomized complete block with
two blocks. Begining 26 Aug. The number of fruit that were on the ground was recorded daily
until all the fruit dropped. Data were collected, pooled for each treatment and regression analyses
were conducted as in the stigma excision experiment. Simple linear regression was used to
define the relationship between number of fruit per tree and average time of drop.
Natural-drop experiment. Trees used in this experiment were 6-year-old ‘Smoothee
Golden Delicious'/M.26, ‘RedChief Delicious’ Mercier strain/M.26, and ‘Commander
York'/Mark maintained using conventional commercial practices. Two trees each of GD and D
and three of Y were used. Fruit under each tree were collected daily until all fruit had abscised.
Data were collected as in the stigma excision experiment and analyzed separately for each
cultivar. Regression analyses were conducted as in the stigma-excision experiment.
Results
Stigma-excision experiment. No multiple regression models explained significantly more
variation than simple linear relationships among the variables: plump seed number, aborted seed
number, fruit weight, and day of year of drop. For D (Fig. 7.1), but not GD and Y (data not
shown) fruit weight increased linearly with increasing numbers of plump seeds. ‘York’ fruit
weight increased linearly with time of drop (Fig. 7.2), but no similar trends existed for D or GD.
There were no significant relationships between aborted or plump seed numbers and day of drop
for any of the cultivars (data not shown).
96
Shade experiment. Shading for 3 d caused all the fruit to abscise (data not presented).
When the 0, 1, and 2 d of shade treatments were pooled, no significant relationships between
date of drop, plump seed number, aborted seed number, and fruit weight were detected (data not
shown). Day of year of drop increased linearly with crop load (Fig. 7.3); however, the range in
date of drop was small (7 d), and the crop loads were light.
Natural drop experiment. No multiple regression models explained significantly more
variation than simple linear relationships among the variables: plump seed number, aborted seed
number, fruit weight, and day of year of drop. Seed number did not affect date of drop for any of
the cultivars. Fruit weight increased linearly with increasing number of plump seeds for D (Fig.
7.4). The periods over which drop occurred for GD, D, and Y were 73, 80, and 94 d,
respectively (Fig. 7.5). Weight of ‘York’ fruit increased linearly with day of year of drop (Fig.
7.5), but the relationship for the other two cultivars, although significant, explained very little
variation.
The distribution of number of plump seeds per fruit was similar for Y and GD but quite
different for D (Fig. 7.6). The mode for all three distributions was eight seeds per fruit. The
proportion of fruit with no plump seeds was much higher in ‘Delicious’ (12%) than in the other
cultivars (<1%) (Fig. 7.6).
Discussion
Seed number per fruit based on individual fruit was not a useful predictor of drop for the
three cultivars in these experiments. The later drop of ‘McIntosh’ fruit with more plump seeds
(Southwick, 1938a) was a mean response of all fruit dropped over multiple periods of 2 to 3 d
97
duration. Our study compared individual fruit dropped on a daily basis and should have provided
greater resolution of differences, yet no relationship was found. Differences in tree age, cultivar,
location, crop load, weather, or other unidentified factors may account for the disparity between
these results. Although Southwick (1938a) reported a linear relationship between seeds per fruit
and date of drop, the relationship was extremely weak (r2 = 0.053).
Late-season fruit growth, after some fruit had dropped, may explain the greater weight of
Y fruit that dropped on later dates. If fruit of all weights have an equal chance of dropping on
any given date, we would expect the fruit that drop later to be larger. Fruit that dropped early
were smaller on Y, but further research is necessary to determine if this is simply a result of the
average increase in fruit weight through the season and the longer period over which they
dropped. The lack of such a trend in the other cultivars suggests that the largest fruit dropped
early and the remaining fruit continued to grow. The D and GD trees also were more vigorous
and had lighter crop loads than did the Y trees (4.2 and 4.6 vs. 8.2 fruit per cm2 trunk cross-
sectional area, respectively). The small Y fruit that dropped early may have been pushed off by
other fruit in the same cluster. ‘York’ has a shorter pedicel than the other two cultivars, and this
could make it more susceptible to such losses.
The increase in D fruit weight with increasing plump seed number (Fig. 7.1) agrees with
reports for ‘Delicious’( Williams, 1977) and ‘Wealthy’ (Murneek and Schowengerhdt, 1935).
Because seed number only explained 25-30% of the variability in fruit weight (Fig. 7.1 and 7.4),
other factors must play a large role in determining fruit size. In the shading experiment, where
crop load varied widely (from 0.96 to 4.04 fruit per cm2 trunk cross-sectional area) , the effect of
98
seed number on weight was not significant. The effect of seed number on fruit size may vary for
different locations, years, crop loads, or cultivars.
Although fruit maturity was not measured, the early drop of fruit from shaded trees may
result from advanced maturity, an effect of reduced crop load. Fruit development and ripening
may have proceeded more rapidly with the lighter crop, and the concomitant increase in ethylene
production could have stimulated drop (Walsh, 1977).
‘Delicious’ matures many fruit with no plump seeds (Williams, 1977), but GD and Y had
very few such fruit (Fig. 7.6). The differences in average number of fruit per tree (727 and 684
for GD and Y vs. 435 for D) may explain the differences in seed distribution. The fruit with no
plump seeds may have been retained on D because of the reduced competition among fruit early
in the season.
In only two of the seven experiments we conducted was there a significant relationship
between fruit weight and seed number, and then seed number only explained 30 and 25% of
variability in weight (Figs. 7.1 and 7.4). Seed number does not appear to be a useful tool for
predicting fruit weight or preharvest drop of the cultivars used in these studies.
99
0
50
100
150
200
250
300 Fr
uit w
t (g)
0 2 4 6 8 10 12 Number of plump seeds
Fig. 7.1. The relationship between fruit weight and number of plump seeds per fruit for
‘Delicious’. Regression line: Y = 156.39 + 6.908X; r2 = 0.303, P# 0.0001, n = 51 (Stigma-
excision expt.).
100
0
50
100
150
200
250 Fr
uit w
t (g)
240 260 280 300 320 340 Day of year of drop
Fig. 7.2. The relationship between fruit weight and day of year of drop for ‘York’. Regression
line: Y = -255.50 + 1.289X; r2 = 0.328, P# 0.0001, n = 84 (Stigma-excision expt.).
101
266
268
270
272
274
Day
of
year
of
drop
50 100 150 200 250 300 350 400 Fruits per tree
Fig. 7.3. Effect of number of fruit per tree on average day of year of drop of ‘Delicious’ fruit.
The relationship is described by Y = 226.09 + 0.0201X; r2 = 0.819, P# 0.0001, n = 6 (Shade
expt.).
102
0
50
100
150
200
250
300
Frui
t wt (
g)
0 2 4 6 8 10 12 Number of plump seeds
Fig. 7.4. Relationship between fruit weight and number of plump seeds per fruit for ‘Delicious’. Regression line: Y = 116.53 +
6.181X; r2 = 0.252, P# 0.0001, n = 878 (Natural-drop expt.).
103
0 40 80
120 160 200 240 280 320
230 250 270 290 310 330
'York'
Day of year of drop
0 40 80
120 160 200 240 280
'Delicious'
0 40 80
120 160 200 240 280 320
'GoldenDelicious'
Fig. 7.5. Relationship between fruit weight and day of year of drop for three apple cultivars.
Regression line for ‘York’: Y = -182.911 + 1.029X; r2 = 0.189, P# 0.0001, n = 2052.
Regression lines not shown for ‘Golden Delicious’ (r2 = 0.0033, P=0.028, n=1459) and
104
‘Delicious’ (r2 = 0.0062, P=0.020, n=872) (Natural-drop expt.).
105
0
5
10
15
20
25
0 1 2 3 4 5 6 7 8 9 10 11 12
'York'
0
5
10
15
20
25
0 1 2 3 4 5 6 7 8 9 10 11 12
'Delicious'
0
5
10
15
20
25
0 1 2 3 4 5 6 7 8 9 10 11 12
'GoldenDelicious'
Fig. 7.6. Frequency distributions for number of plump seeds per fruit for ‘Golden Delicious’ (n =
1455.), ‘Delicious’ (n = 870), and ‘York’ (n = 2017) trees (Natural-drop expt.).
106
Literature Cited
Dennis, F.G. 1967. Apple fruit set: Evidence for a specific role of seeds. Science 156:71-73.
Luckwill, L.C. 1953. Studies of fruit development in relation to plant hormones I. Hormone
production by the developing apple seed in relation to fruit drop. J. Hort. Sci. 28:14-24.
Murneek, A.E and G.C. Schowengerhdt. 1935. A study of the relation of size of apples to
number of seeds and weight of spur leaves. Proc. Amer. Soc. Hort. Sci. 33:4-6.
Southwick, L. 1938a. Pre-harvest drop of the McIntosh apple. MS Thesis. Massachusetts State
College, Amherst, Mass.
Southwick, L. 1938b. Relation of seeds to pre-harvest McIntosh drop. Proc. Amer. Soc. Hort.
Sci. 36:410-412.
Walsh, C.S. 1977. The relationship between endogenous ethylene and abscission of mature apple
fruits. J. Amer. Soc. Hort. Sci. 102:615-619.
Williams, M.W. 1977. Adverse weather and fruit thinning chemicals can affect seed content and
size of ‘Red Delicious’ apples - What can growers do about it? Proc. Washington State Hort.
Assoc. 73:157-161.
107
Chapter Eight
Time Course and Effect of Environmental Factors on Preharvest Apple Fruit Drop.
Abstract. Apple trees often drop much of their crop before optimum harvest time. The
distribution of drop of apple fruit is poorly defined and ecophysiological processes associated
with drop are poorly characterized. Three cultivars of apple, ‘Smoothee Golden Delicious' (GD),
‘RedChief Delicious' (D), and ‘Commander York'(Y) were studied for three years to 1) estimate
variance due to year and cultivar, 2) identify environmental factors associated with drop, and 3)
test whether rewetting by rain after a drier period induced drop. Time of drop of all fruit on the
tree was recorded. Variance of average day of drop from year to year was 40.1 while variance
among cultivars within a year was 51.8 and variance from tree to tree within each cultivar within
each year was only 18.6. Multiple regression modeling of time of drop for each cultivar within
each year revealed that much of the variability in time of drop was due to factors other than the
weather events modeled. The best models developed explained only 8% to 35% of the variability
in time of drop. The most important weather factors were daily minimum temperatures and
precipitation. Rain events of greater than 0.2 inches following a drier period caused increased
drop in one of three years.
Many factors may participate in controlling the fruit abscission process and the time of
fruit drop. Environmental factors like wind and weather patterns (both daily and seasonally) can
be expected to play a major role in determining when fruit abscise, but predicting fruit drop
would require much more knowledge including weather effects on fruit drop. Environmental
factors may affect the process both directly and indirectly through other processes with which
108
abscission may be coordinated. Normal developmental processes for the annual cycle of apple
tree growth like maturation and ripening of the fruit, and many photoperiod-induced changes
during the autumn appear at least temporally associated with abscission and may be regulated in
conjunction with abscission.
Available information regarding the effect of weather on preharvest apple fruit drop is
limited. Southwick (1938) reviewed drop records over 16 years and found no significant
correlation between weather conditions and preharvest fruit drop of ‘McIntosh’(1938). The
retrospective evaluation of drop records where data were only collected at intervals may not have
provided sufficient precision to accurately evaluate weather effects. A study was therefore
initiated to identify weather factors associated with daily fruit drop prospectively.
The objectives of this study were to: 1) characterize the time profile of drop for all the
fruit on the tree, 2) identify relationships between weather factors and daily fruit drop, and 3)
investigate the hypothesis that rain events following a period of dry weather induces drop.
Materials and Methods
Trees used in this experiment were six, and seven, and eight-year-old ‘Smoothee Golden
Delicious'/M26 (GD), ‘RedChief Delicious'/M26 (D), and ‘Commander York'/Mark (Y) growing
at the Virginia Tech College of Agriculture and Life Sciences Kentland farm near Blacksburg,
Va. In each year two or three trees of each cultivar were included in this data set. The orchard
floor under the trees was cleaned and each day the dropped fruit were counted and removed until
all fruit on the tree had dropped. Daily weather records from the farm’s weather station were used
to provide precipitation and temperature data.
To describe the time of year of drop the mean day of the year of drop of all the fruit on a
109
tree was calculated for each year and cultivar. Graphical analysis of plots of percentage of crop
dropped across days were used to visualize the distribution of drop for each year and cultivar. To
estimate variance due to years, cultivars, and trees a random effects model was fit with cultivar
nested within year and tree nested within cultivar. The weighted average day of drop for each
tree was used as the response variable. Maximum likelihood estimation of the variance
components was performed using the MIXED procedure of the SAS System. Our purpose was to
identify the contributions of these effects to drop as if the years, cultivars and trees in our
experiment are randomly sampled from a population of possible years, cultivars and trees. Year
and tree can easily be seen as random effects, but for cultivar we are relaxing the definition of a
random effect to adapt the method to our objective.
The percentage of the total crop that fell on each day was used as the response variable
for multiple regression modeling with the goal of identifying important environmental factors
associated with drop from among our measured weather variables. Variable selection was the
primary objective of this analysis so all possible regressions were fitted and sorted by R2 values
to begin the selection process. The categorical predictors year and cultivar were considered
orthogonal because they were crossed factors by design. The initial all possible regressions
analyses included indicator variables for year and cultivar, but because they were both important
explanatory variables a choice had to be made whether to divide the data set up by year and
cultivar or attempt to develop the model using one grand model with many more candidate
regressors to break up the effects of the year and cultivar and their interactions with the other
regressors. Although separate models are less satisfying for making sweeping interpretations, the
data set was divided up by year and cultivar to simplify the modeling and to improve the
interpretability of the final models.
110
Continuous regressor variables were all centered to reduce the effects of collinearity
between the main effects and their interactions. All the continuous regressors (maximum
temperature, minimum temperature, precipitation, average minimum and maximum temperature
during the seven previous days, and average daily precipitation over the previous seven days) and
their two-way interactions were included in the pool of candidate regressors. All possible
regressions were fit for each year by cultivar combination.
Plots of R2, and MallowsCP across number of regressors in each model were used as an aid
to select a model size that included enough regressors to approach minimum prediction error
without overspecifying the model (avoid P >> Cp). Models were constrained to be hierarchically
well formulated (main effects participating in an interaction were always included in the model).
Diagnostics were performed to assess heteroskedasticity and influence of individual
observations (Studentized residuals, leave-one-out studentized residuals and PRESS residuals).
Inspection of standardized residual plots suggested a multiplicative error structure so the
percentage of crop dropped was log-transformed to stabilize variances. Collinearity among
regressors was assessed using scatter plot matrices, and variance inflation factors.
A separate analysis was performed using a completely prespecified model to test the
hypothesis that substantial rain events following a drier period would induce drop. This analysis
was performed for each year and cultivar combination separately. The response analyzed was
percentage of crop dropped. Data were log transformed to stabilize variances and back-
transformed for presentation. For this analysis a substantial rain was defined as more than 5 mm
of rain on that day. An indicator variable was created which classified all days as either having
substantial rain or not. The regression model contained the indicator variable, the average daily
precipitation over the previous seven days and their interaction. The GLM procedure of the SAS
111
System was used to perform a fixed effects analysis of covariance.
Results and Discussion
The three cultivars examined in this study were selected because of their economic
importance to the region and because they represent a range of severity of preharvest drop
behavior. GD was expected to exhibit little preharvest drop while D and Y are both prone to
more severe preharvest drop.
Variation among cultivars is well known, but even cultivars that have a tendency to drop
fruit heavily vary substantially from year to year. Variance of average day of drop from year to
year was 40.1 while variance among cultivars within a year was 51.8. By comparison the
variability from tree to tree within each cultivar within each year was only 18.6. In these years,
with these cultivars there was comparatively little variation among trees but year and cultivar
both contributed substantially to the variation in day of drop. Because years vary widely,
predicting time of drop in any given year will only be possible if we can identify those factors
responsible for this variation.
Y had the latest drop in each year followed by GD and D respectively (Table 8.1). Y
matures later than GD which matures later than D and this may be responsible for much of their
variability in day of year of drop.
The trees in these experiments did not exhibit severe preharvest drop in any year yet the
variability from year to year of drop is evident (Table 8.1). Observation of the drop profile (Fig.
8.1) reveals that the drop was much later in 1997 than in either 1996 or 1995. The distribution of
drop appears different among the cultivars. Of these three cultivars GD appears to drop fruit
during a more limited period while drop of Y and especially D is spread over a wider time. The
112
cultivars with more serious preharvest drop problems tend to drop their fruit over a longer period
of time.
Multiple regression models for each year and cultivar were fit beginning with all possible
regressions. The proportion of total variation in drop explained (R2) by models with different
numbers of regressors was inspected as a guide to the number of regressors important in
explaining drop (Fig. 8.2). There were large differences in the proportion of variation explainable
by regression modeling. This indicates that overall the impact of weather factors is less in some
year and cultivar combinations than in others. In no case was there substantial additional
explanation of drop by including more than six or seven regressors. At this stage, even without
selecting the optimum models, it is evident that factors other than the weather factors included in
our pool of candidate regressors are responsible for much of the variation in drop.
To validate the sizes suggested by the increases in R2, Mallows Cp (a measure of the error
of prediction for a regression model) of the 10 best fit models of each size was plotted (Fig. 8.3).
Regression models are likely overspecified when the number of regressors is larger than Mallows
Cp (Myers, 1990), so a reference line was added to the plot and the number of regressors
suggested for each year by cultivar combination was estimated by the smallest models that were
near their own values for Mallows Cp. Models with approximately this number of regressors
were then developed for each year by cultivar combination. For example, 6 regressors were
included in the model for D in 1995.
The all possible regressions provided information about which regressors were most
important in explaining drop. Many times an interaction term was identified as important so
these were included in the model and the main effects participating in the interaction were forced
into the model to maintain interpretability (Myers, 1990). When the same regressors appear in
113
several of the best fitting models of a given size those regressors were considered more important
to include in the final model. Comparisons of models with similar explanatory value were made
based on their simplicity and their consistency with models having somewhat smaller or larger
numbers of regressors.
For each year by cultivar combination a final model was developed and coefficients for
each regressor were estimated. The final models explained between 8% and 35% of the total
variability in percentage of crop dropped on a day (Table 8.2). As seen in the unreduced analysis,
much of the variability in drop remained unexplained by the weather factors.
Some regressors were more consistently associated with drop, especially inches of
precipitation, minimum temperature on day of drop, and average minimum temperature over the
seven days preceding drop. A simple explanation of the effect of minimum temperature is that
more drop occurs later in the season when temperatures are lower. Precipitation is potentially
associated with drop because of physiological effects, but much of its effect may be due to the
presence of confounding factors not included in these analyses, especially wind occurring with
the precipitation. Unfortunately accurate wind data were unavailable for our site, but further
investigation of its effect is certainly warranted.
The interaction of precipitation with average minimum temperature over the previous
seven days is in the final models for all cultivars in 1995 and 1997, but in none of them in 1996
(Table 8.2). To visually explore the nature of this interaction a spline-smoothed surface of the
predicted percentage of crop dropped based on the full model was made for each year and
cultivar combination (Fig. 8.4). In 1995 with all three cultivars drop decreased as precipitation
increased when the temperature had been lower over the preceding seven days, but not when the
temperature had been higher. In 1997 the effect was very different, drop was greater when daily
114
minimum temperatures had been lower over the previous seven days, but this effect was reduced
by higher precipitation. The later time of drop for all cultivars in 1997 (Table 8.1) may have
caused the difference in these results, as much of the crop was retained until temperatures were
lower in 1997. The year to year variation in drop is not only a difference in when the fruit drop,
but also in what factors are responsible for causing drop in a year.
Field observations had suggested the hypothesis that rain following a drier period would
induce significant drop. There was a significant interaction between substantial rain and average
rainfall over the preceding seven day period for all cultivars in 1995, but not in the other two
years (Fig. 8.5). The nature of the interaction was that more drop occurred with rain following a
drier period. Weather conditions in the three years of these studies only provide a preliminary
evaluation of this effect, but the effect was visible in one of the three years. In some years
substantial rain events following periods with less rain may cause serious drop.
115
Table 8.1. Mean (± standard deviation) day of the year of drop of apple fruit from six, seven, and
eight-year old trees of ‘Smoothee Golden Delicious'/M26 (GD), ‘RedChief Delicious'/M26 (D),
and ‘Commander York'/Mark (Y).
Cultivar
Year D GD Y
1995 278.6 (±13.2) 283.3 (±11.1) 288.9 (±19.4)
1996 269.3 (±18.0) 284.9 (±16.3) 295.9 (±13.6)
1997 297.3 (±12.5) 301.1 (± 7.8) 292.3 (±13.8)
116
Table 8.2. Final multiple regression models and R2 for each year and cultivar. Cultivars are
‘RedChief Delicious’ (D), ‘Smoothee Golden Delicious’ (GD), and ‘Commander York’ (Y). The
modeled response is the natural logarithm of the percentage of total crop dropped on a dayz.
Year CV R2 Regression model
1995 D 0.25 Lpct_crop = 0.085 + 0.62*Prcp + 0.30*Mintcum7 - 0.38*
PrcpxMintcum7
GD 0.08 Lpct_crop = 0.008 + 0.89*Prcp + 0.11*Mintcum7 -
0.64*PrcpxMintcum7
Y 0.08 Lpct_crop = 0.16 + 1.11*Prcp - 0.007*Mintemp + 0.036*Mintcum7 -
0.35*PrcpxMintcum7
1996 D 0.12 Lpct_crop = 0.32 - 0.004*Mintemp + 0.006*Maxtemp -
0.004*MintempxMaxtemp
GD 0.27 Lpct_crop = 0.31 - 0.49*Prcp - 0.04*Mintemp + 0.03*Maxtemp -
0.006*MintempxMaxtemp - 0.12*Mintcum7
Y 0.21 Lpct_crop = 0.29 - 0.16*Prcp - 0.03*Mintemp - 0.09*Mintcum7 -
2.69*Prcpcum7 - 4.69*PrcpxPrcpcum7 - 0.03*MintempxMintcum7
1997 D 0.30 Lpct_crop = 0.07 + 0.86*Prcp - 0.05*Mintemp + 0.004*Maxtemp -
0.22*Mintcum7 + 0.006*MaxtempxMintcum7 -
0.64*PrcpxMintcum7
GD 0.35 Lpct_crop = -0.16 + 1.49*Prcp - 0.04*Mintemp - 0.20*Mintcum7 +
0.01*MintempxMintcum7 - 0.02*PrcpxMintemp -
1.27*PrcpxMintcum7
117
Y 0.20 Lpct_crop = 0.21 + 0.80*Prcp - 0.06*Mintemp -
0.09*PrcpxMintcum7
z Regressor abbreviations: Mintemp = Minimum temperature on day of drop (0C), Maxtemp =
Maximum temperature on day of drop (0C), Precip=Inches of precipitation on day of drop,
Mintcum7=Average minimum temperature over the seven days preceding drop (0C), Prcpcum7 =
Average inches of precipitation over the seven days preceding drop.
122
temperature increases from left to right.
124
interaction between effect of rainfall greater than 5 mm and effect of average rainfall over the
seven days preceding drop.
125
Literature Cited
Myers. R. H. 1990. Classical and modern regression with application. PWS-Kent, Boston.
Southwick, L. 1938. Pre-harvest drop of the McIntosh apple. Master's Thesis. Mass. State
College, Amherst
126
General Discussion
To reduce drop losses we could hypothetically build protective structures around the tree
to protect it from buffeting winds and predators. Or, given time, we might introduce genetic
modifications to reduce abscission (Sato, et al., 2004). We could also continue to exploit plant
growth regulating substances to reduce preharvest fruit abscission. These approaches all have
limitations. Government regulation makes PGRs undependable tools in the long run. Genetic
modification provides many long-term possibilities, but is expensive to perform and requires
much time to test and become adopted. Growing trees in protective structures is currently too
expensive for a crop like apple. Preferably we could modify training, pruning, thinning or
harvesting practices to minimize loss from preharvest fruit abscission. The studies in this work
were intended to identify sources of variability in drop that could be used to develop improved
cultural practices for limiting preharvest drop.
One of the greatest difficulties in studying fruit drop is the short period each year when
drop occurs. Inferences from model systems may not always behave similarly to the actual
orchard situation, but they are still valuable for studying specific physiological aspects of
abscission. The cut fruit system can be used in the orchard and avoids the artificiality of an
explant but, nevertheless creates an artificial situation.
The large wound created by cutting likely creates a stronger ethylene signal than ever
experienced by the fruits about which we wish to make inferences. Such a gross effect likely
elicits a quantitatively and qualitatively different response than otherwise produced by the tree
during typical fruit drop. The wounds could be inducing a response typical of responses to
wounds from predators, etc. Notwithstanding these limitations, the cut fruit system has been
found to be effective for predicting efficacy of PGRs as drop delaying and inhibiting tools
127
(Marini and Byers, 1988).
Our experiments demonstrated the interaction of NAA with AVG on delay of drop with
cut fruit, suggesting that because their effects are more than additive in our system they should be
more than additive in the orchard. And this has been verified in commercial orchards (personal
communication with R. Marini). Combined applications of NAA and AVG could be tried on an
experimental basis. By this route the orchard manager can hopefully achieve effective drop
control even with cultivars and in blocks that are prone to drop. Hopefully this will also reduce
the cost of the PGRs making it feasible for use where its cost might otherwise be prohibitive.
The period of reduced sensitivity to abscission induction may provide an opportunity for
manipulating plant development. If we assume that there is an ethylene-inducible, ethylene
binding protein receptor functioning in regulating fruit abscission, then perhaps even when little
ethylene is produced it causes the production of ethylene receptors. A hypothesis to consider
when designing future experiments is that AVG applied during the later part of the period of
reduced sensitivity to abscission induction will prevent the autocatalytic, feed-forward, increase
in ethylene production and sensitivity. This may work by preventing the de novo synthesis of
ethylene binding proteins; therefore preventing sensitivity to ethylene from increasing.
It would be desirable to perform new experiments to determine if the tandem application
of NAA and AVG can be timed to take advantage of the period of change in sensitivity to
abscission induction. A range of timings of AVG applications followed by a typical application
of NAA would be valuable to determine the optimum timing for combined effect. We propose as
the target for the AVG the changes occurring during the resumption of sensitivity to abscission
induction. Theses changes include resensitizing abscission to ethylene. During the time of
qualitative change in sensitivity it is expected that small increases in ethylene production could
128
translate into large differences later.
The finding that fruit which drop early are riper than those on the tree at that time
suggests that uniform ripening among the fruit on the tree would reduce preharvest drop losses.
Training and thinning practices to increase uniformity among fruit require further investigation to
determine if they can reduce preharvest drop. The earlier drop of the highest fruit in the tree
canopy also suggests that differences in ripening could lead to the differences in time of drop
among fruit. With older, larger trees than those used in these studies this effect may be more
pronounced.
Although these studies failed to detect any effect of the attachment of the fruit on time of
drop, there may be effects that we failed to investigate. The structural differences among fruit at
the anatomical level may play an important role in determining which fruit fall first.
Excision of stigmata resulted in reduced fruit set and retention, but when even one stigma
remained many were able to fruit set. This indicates that incomplete pollination and fertilization
is sometimes adequate for acceptable fruit set in these cultivars. In situations with more intense
competition among fruit at time of set, more seeds may be a larger advantage to setting. The
reduced number of seeds per fruit obtained by stigma excision did not translate into earlier fruit
drop though. Variation in seed number among fruit was not directly associated with earlier drop.
In other years seeds may play a larger role in determining fruit size and this may lead to an effect
on early drop, but these connections would be difficult to elucidate given the longer time fruit
have to grow when they drop later (Byers and Eno, 2002).
Unfortunately, these studies were not performed so as to allow investigation of diurnal
patterns of drop. Observations of diurnal drop patterns may provide useful insight into the drop
process in response to processes during the course of each day. Water status of the whole tree,
129
especially during rehydration, may affect swelling of AZ cells. The short-term changes during
the day combining more turgid, swollen cells with breezes or wind gusts may play an important
role in determining the time of drop. Other diurnal variation may explain key concepts in our
understanding that would likely have been missed by all previous study.
The increased drop following rain events may be partially caused by swelling of cells in
the AZ of the now well hydrated trees. The isotropic swelling of AZ cells seen under the
microscope may be generating forces that break strands of undegraded vascular cells. The
mechanical force needed to break the fruit loose is likely delivered by the wind and rain during a
rain event, but the swollen cells of the AZ may also act as superior fulcrums for the breaking
action. A question remaining to be answered is whether careful use of irrigation water could be
used as a cultural tool for reducing preharvest drop. Some questions about effects of water
relations on drop may be answerable under arid orchard conditions with irrigation.
Deciding when to harvest is driven by many factors, from fruit color development to
labor availability. Preharvest drop also figures into the decision of when to harvest. Our
observations of increased drop associated with severe rain and wind storms is probably
understood by many growers. And we think it makes sense that an impending severe storm may
be a reason to harvest slightly earlier. When possible, harvesting before a storm rather than after
it could save substantial fruit.
Is it a better idea to allow the fruit to drop than to bother with stop-drop practices? For
those blocks of trees that drop heavily and for cultivars with significant drop problems the losses
from drop will justify control measures in most years. Whereas many times the decision will
need to be made when the loss from drop is very hard to predict. Our best estimate of the
potential loss and the cost of eliminating the stop-drop sprays has the clear benefit of reducing
130
the cost of production. An important effect of delaying harvest even though fruit are falling is
that the fruit continue to grow during such time. Any increase in yield obtained will partially
offset the fruit lost to drop (Byers and Eno, 2002). We have seen that the fruit that drop early are
riper than those on the tree at the same time. Therefore, allowing fruit to drop will eliminate
proportionately more of the most mature fruit from the harvest. This loss of fruit might be
partially offset if the uniformity of the harvest is increased.
131
Literature Cited
Byers, R. E., and D. R. Eno. 2002. Harvest date influences fruit size and yield of 'York' and
'Golden Delicious' apple trees. J. Tree Fruit Prod. 2002. 3(1):63-79.
Marini, R. P and R. Byers. 1988. Methods for evaluating chemical inhibitors of apple abscission.
HortScience. 23(5):849-851.
Sato, T., T. Kudo, T. Akada, Y. Wakasa, M. Niizeki, and T. Harada. 2004. Allelotype of a
ripening-specific 1-aminocyclopropane-1-carboxylate synthase gene defines the rate of fruit drop
in apple. J. Am. Soc. Hort. Sci.129(1):32-36.
132
Vita
Daniel Lee Ward
Graduated from Sandy Valley High School, Magnolia, Ohio
Bachelor of Horticulture. The Ohio State University, Columbus, Ohio.
Fruit and Vegetable Production Option.
Master of Science, Plant and Soil Science. Southern Illinois University, Carbondale, Illinois.
Thesis title:Reducing flower bud density of 'Redkist' peach with GA3.
Doctor of Philosophy, Horticulture. Virginia Polytechnic Institute and State University,
Blacksburg, Virginia.
Dissertation title: Factors affecting preharvest fruit drop of apple.