This article appeared in a journal published by Elsevier. The attached
copy is furnished to the author for internal non-commercial research
and education use, including for instruction at the authors institution
and sharing with colleagues.
Other uses, including reproduction and distribution, or selling or
licensing copies, or posting to personal, institutional or third party
websites are prohibited.
In most cases authors are permitted to post their version of the
article (e.g. in Word or Tex form) to their personal website or
institutional repository. Authors requiring further information
regarding Elsevier’s archiving and manuscript policies are
encouraged to visit:
Author's personal copy
Timing of fruit removal affects concurrent vegetative growth and subsequent
return bloom and yield in olive (Olea europaea L.)
Arnon Daga,*, Amnon Bustana, Avishai Avnia,b, Isaac Tziporia, Shimon Laveeb, Joseph Riovb
aGilat Research Centre, Agricultural Research Organization, Ministry of Agriculture, Mobile Post Negev 85280, Israel
bThe Kennedy-Leigh Centre for Horticultural Research, The Robert H. Smith Institute of Plant Sciences and Genetics in Agriculture, The Hebrew University of Jerusalem,
Rehovot 76100, Israel
fruit production. Being an industry-dependent commodity, the
economic impact of biennial bearing, particularly in oil olives, is
highly significant (Lavee, 2006). Alternate bearing is intimately
related to the basic processes of fruiting, such as flower bud
differentiation, fruit set and abscission, and fruit growth (Gold-
schmidt, 2005). In olive, as in many other fruit trees that display
biennial bearing, the syndrome is first expressed by the intensity of
On-year, and conversely—an Off-year is tailed by profuse flowering
(Monselise and Goldschmidt, 1982; Lavee, 1996; Goldschmidt,
2005). Thus, research into alternate bearing in this species should
focus on the early phases of the reproductive process.
Olive fruits develop on inflorescences arising from buds borne
on shoots grown in the previous year (Lavee, 1996). Sanz-Corte ´s et
al. (2002) suggested that floral induction occurs in the summer, 7–
8 weeks after full bloom, at about the time of pit hardening
(endocarp sclerification) in the concurrent season’s fruit. Indeed, a
more recent work (Andreini et al., 2008) could already distinguish
between On and Off axillary buds in July, close to pit hardening. At
that time, accumulation of the cytokinin zeatin was observed only
both considered early indicators of floral initiation. Nevertheless,
olive shoots often grow throughout the summer, so that the buds
developing in the leaf axils along the shoot are of diverse ages,
some emerging after pit hardening. Nevertheless, all of the buds on
well-lignified parts of the shoot can potentially differentiate to
form inflorescences (Lavee, 2006).
Floral bud development in olive is governed by plant growth
substances and various secondary metabolites delivered from
other plant organs (Badr et al., 1970; Ferna ´ndez-Escobar et al.,
1992; Lavee, 1996; Fabbri and Benelli, 2000; Baktir et al., 2004;
Ulger et al., 2004), and is also strongly affected by winter
temperature regime (Martin, 1989; Rallo and Martin, 1991;
Rallo et al., 1994; Lavee, 1996; Orlandi et al., 2004). Cuevas et al.
(1994) showed that heavy fruit load impairs the floral quality of
the following bloom, presumably through incomplete bud
differentiation. Experiments involving deliberate seed abortion
(Stutte and Martin, 1986) or removal of young developing fruit
(Ryan et al., 2003) demonstrated the negative influence of
Scientia Horticulturae 123 (2010) 469–472
A R T I C L EI N F O
Received 19 April 2009
Received in revised form 11 October 2009
Accepted 23 November 2009
A B S T R A C T
Olive (Olea europaea) demonstrates a high tendency toward alternate fruit production, with significant
negative consequences on the industry. Fruit load is one of the main cause-and-effect factors in the
phenomenon of biennial bearing, often disrupting the balance between reproductive and vegetative
processes. The objectives of the present study were to identify the time range during which heavy fruit
load reversibly interrupts the reproductive processes of the following year. The linkage between timing
of fruit removal, vegetative growth, return bloom, and fruit yield was studied. Complete fruit removal in
cv. Coratina until about 120 days after full bloom (August 15) caused an immediate resumption of
vegetative growth. The new shoots grew to twice the length of those on trees that underwent later fruit
removal. Moreover, a full return bloom, corresponding with high subsequent yields, was obtained by
earlyfruitremoval,whilepoorornobloom developedonlate-defruited orcontroltrees.Thus,thecritical
time to affect flowering and subsequent fruiting in the following year by fruit thinning occurs in olive
trees even weeks after pit hardening—much later than previously suggested. Furthermore, the data
indicate that flowering-site limitation, due to insufficient or immature vegetative growth during the On-
year, is a primary factor inducing alternate bearing in olive.
? 2009 Elsevier B.V. All rights reserved.
* Corresponding author at: Gilat Research Centre, Agricultural Research
Organization, Ministry of Agriculture, Mobile Post Negev 85280, Israel.
Tel.: +972 8 9928630; fax: +972 8 9926485.
E-mail address: firstname.lastname@example.org (A. Dag).
Contents lists available at ScienceDirect
journal homepage: www.elsevier.com/locate/scihorti
0304-4238/$ – see front matter ? 2009 Elsevier B.V. All rights reserved.
Author's personal copy
phenolic compounds induced by the developing seed or fruit on
floral bud differentiation. A high positive correlation was shown
between fruit load and the level of chlorogenic acid in the leaves
in mid-summer. Injection of chlorogenic acid into scaffolds
during winter, but not in the following early spring (Lavee et al.,
1986), reduced flower bud differentiation by more than 50%
(Lavee, 2006). This bulk of evidence suggests that the concurrent
developing fruit directly inhibits the metabolism leading to
reproductive induction and differentiation of the buds for the
potential yield in the following year.
Adequate development of new shoots is essential for the
successive reproductive process, as these serve as the platform for
new buds. As a consequence of heavy fruit load, the vegetative
2005; Lavee, 2006). Thus, the vicious circle of biennial bearing is
undoubtedly the consequence of an imbalance between the
vegetative and reproductive phases of the tree. The balance
between these two phases may often be interrupted by environ-
mental factors, such as unfulfilled winter chilling requirements
(Rallo and Martin, 1991; Rallo et al., 1994) or spring heat waves
(Lavee, 2006), which radically reduce fruit load, thereby triggering
the circle by inducing an Off-year.
Once initiated, the main method of breaking biennial bearing
is On-year fruit thinning (Dag et al., 2009). Strong interactions
between developing fruit, shoot vigor, and the induction and
differentiation of floral buds have been described in olive trees
with regard to biennial bearing (Lavee, 2006). However, better
resolution of the time course of the processes involved, and
particularly the identification of points of no return in the
subsequent reproductive development, is still needed. In the
present study, complete fruit removal at different times during
the season was employed to identify the critical time at which
heavy fruit load prevents reproductive development in the
following year, and to study the linkage between fruit removal,
vegetative growth, return bloom, and fruit yield in olive trees
during the season, in the context of biennial bearing.
2. Materials and methods
2.1. Experimental site and design
The experiment was conducted in a commercial olive (cv.
Coratina) orchard planted in 2002 (at 7.0 m ? 3.5 m spacing) in an
arid area near Kibbutz Revivim in the Negev highland desert of
Israel. The mean annual precipitation is <100 mm between
November and February and it is unpredictable; the orchards
are drip-irrigated throughout the year to reach a total of 900 mm.
Fertilizers are supplied continuously through the irrigation water
at 200, 30, and 300 kg ha?1year?1of N, P and K, respectively.
Forty-eight uniform heavily producing trees were selected and
tagged on June 6, 2006, after the final fruit load had been
determined. The trees were randomly assigned to the different
treatments, six trees (replicates) per treatment (date of fruit
removal). Accordingly, all fruits were manually removed from the
trees on June 6, July 12, August 15, October 17, November 8, and
December 6,2006 (days of year [DOY]: 157, 193, 227, 290, 312, and
346, respectively), and January 11, 2007 (DOY: 11). The eighth
treatment consisted of trees whose fruits were not removed,
serving as a control.
On June6, 2006, four representativeshoots pertree, twooneach
trees, and their growth parameters (shoot elongation, internode
length, and branching) were followed once a month. Sunlight
measured once, withthe device locatedhorizontallyatthe centerof
the tree, about 0.4 m above the ground, where the trunk splits to
main branches Each measurement inside the canopy was preceded
by an above-canopy one. All measurements were normalized to the
mean light interception of the 6 control trees thus determining the
relative canopy density (RCD).
The intensity of the return bloom was evaluated on April 15,
2007, using a blind test in which two external surveyors
independently ranked each experimental tree from 0 (no bloom)
to 5 (heavy bloom). The subsequent year’s fruits were harvested
from individual trees on November 4–6, 2007 onto nets using
mechanical combs, gathered, and weighed. A sample of 100 fruits
was taken from each tree to determine the average fruit weight,
and number of fruits per tree was calculated.
Data were analyzed using JMP 7.0 software (SAS Institute).
Effect of treatments on yield parameters and RCD were analyzed
using a one-way ANOVA model (Tukey–Kramer multiple compar-
3. Results and discussion
Developing olive fruit has been reported to have a marked
negative effect on the subsequent developmental processes that
Stutte and Martin, 1986; Cuevas et al., 1994; Ryan et al., 2003;
Lavee, 2006). Most previous studies have dealt mainly with floral
bud induction and differentiation. Evidently, however, the
number of new shoots developing in a year during the growing
the number of available buds for floral induction. The number of
such buds is the primary determinant of reproductive capacity in
the following year.
In the present study, complete manual fruit removal was
applied once a month during the season in young ‘Coratina’ trees.
The effect of fruit removal on the shoot growth was significant:
until about 120 days after full bloom (DAFB) (late August), the
earlier the fruit was removed, the more vigorous the shoot growth
as soon as this restraint was lifted by fruit removal, vegetative
Fig. 1. EffectofmanualfruitremovalondifferentdatesafterfullbloomduringanOn-
year, starting on June 6, 2006 (50 days after full bloom), on shoot elongation of olive
trees (cv. Coratina). Values are means of 24 shoots (four per tree). Bars indicate ?SE.
A. Dag et al./Scientia Horticulturae 123 (2010) 469–472
Author's personal copy
growth resumed, resulting in approximately twice the shoot-
elongation increment compared to trees defruited later in the
season. In all treatments, shoot growth almost ceased at about 180
DAFB. After that time, fruit removal was too late to enable
significant reinduction of vegetative growth. Relative canopy
density (RCD) values measured in the following spring, prior to
bloom (Fig. 2) closely correlated with the shoot growth measure-
ments during the season, confirming the overall influence of fruit
removal on the vegetative growth.
Although renewed vegetative growth following fruit thinning
(Ferna ´ndez-Escobar et al., 1992) or embryo killing (Stutte and
Martin, 1986) has been demonstrated, the effectiveness of
reducing fruit number on vegetative growth was limited to the
period until pit hardening, less than 2 months after full bloom.
Our data show that the negative impact of fruit load on
concurrent vegetative growth can be reversed considerably later
by fruit removal, up to 4 months after full bloom (Fig. 1). Since
no significant influence has been observed on the length of the
new internodes (data not shown), it is presumed that the
measured shoot elongation has brought about to a significant
increment in the number of internodes, and most likely—in the
number of new buds per tree. Thus we show here, that the
potential for flowering and fruiting in the following year can be
completely restored when fruit in the On-year is removed, up to
It is difficult to distinguish between the inhibitory effect of
developing fruit on vegetative growth and their direct influence on
floral bud initiation and development. Several authors have
suggested that flower bud induction occurs in early July, 7–8
weeks after full bloom, at around the time of pit hardening (Sanz-
Corte ´s et al., 2002; Andreini et al., 2008). In the present study,
fruitlet removal until 120 DAFB provided high blooming scores in
the subsequent spring; however, fruit removal at any later date
had only a minor effect on the flowering rates, which were
nevertheless significantlyhigher than the null bloom of the control
trees (Fig. 3). These results demonstrate clearly that the buds on
the newly developing shoots (induced to grow by fruit removal up
to 120 DAFB) had the capacity to differentiate, 1 or 2 months later
than assumed before.
flower bud induction and the developmental stage of the shoots
bearing the buds and that of the buds themselves. Shoot growth
and development is a prolonged process hence the population of
new axillary buds on a tree at any given time is rather
heterogeneous with respect to their responsiveness to inducive
or inhibitory effects. It follows that floral induction of the whole
population of new buds is likely to be scattered throughout the
season, even though this is considered a finite event for the
individual bud (Lavee, 2006). The bud population is subject to
various regulatory effects, among which may be age, location, and
vicinity of an individual bud to developing fruit. As a rule, the
earlier the fruit removal the more new buds are released from the
inhibitory influence. Nevertheless, a fair amount of buds remain
responsive to inducive factors and may differentiate even if fruit
removal takes place relatively late (Fig. 3). It could therefore be
concluded that alternate bearing is controlled indirectly by the
number of buds, but also directly by the readiness of each
individual bud to differentiate at a given time. It is also possible
that a second notable period of induction, which may restore or
strengthen the effect of earlier ones, occurs during the autumn
(Troncoso et al., 2010).
Adequate number of responsive buds and their subsequent
differentiation are indeed prerequisites but they cannot guarantee
fruit yield. In our study, the fruit yields corresponded to the extent
of the bloom: high yields (>50 kg tree?1) were obtained in trees
that had undergone fruit removal until 120 DAFB in the previous
season (Table 1). In contrast, fruit removal after this time gave rise
to much lower yield levels (7–13 kg tree?1). It should be noted that
a slight increase in flowering and subsequent (though negligible)
fruit production, whereas no fruit was obtained in the control trees
(no fruit removal at all). The difference in fruit load between the
Fig. 2. Effectofmanualfruitremovalatdifferent timesafterfullbloominanOn-year,
starting on June 6, 2006 (50 days after full bloom—DAFB), on the relative canopy
density (RCD) of olive trees (cv. Coratina) in the following spring (March 12, 2007).
Values are means of six trees. Bars indicate ?SE. Means headed by different letters are
significantly different at P ? 0.05 (Tukey–Kramer multiple comparisons test).
Fig. 3. Effect of manual fruit removal at different times after full bloom in an On-
year, starting on June 6, 2006 (50 days after full bloom—DAFB), on the following
bloomofolive trees(cv.Coratina), ranked (0—nobloomto5—heavy bloom) onApril
15, 2007. Values are means of six trees ? SE.
Effect of timing of fruit removal (in brackets: day of year) in cv. Coratina on fruit-
yield parameters in the following year.
Date of fruit removalFruit load
June 6, 2006 (157)
July 12, 2006 (193)
August 15, 2006 (227)
October 17, 2006 (290)
November 8, 2006 (312)
December 6, 2006 (346)
January 11, 2007 (11)
Within each column data followed by different letters are significantly different
according to the Tukey–Kramer multiple comparison test (P<0.05).
A. Dag et al./Scientia Horticulturae 123 (2010) 469–472
Author's personal copy Download full-text
two groups (fruit removal before and after 120 DAFB) had a clear
effect on fruit size, which was significantly smaller in the early
defruited, high-yielding trees (Table 1).
To conclude, the requisite existence of new, mature and
therefore responsivebudsseems tobe the mainfactor determining
the return bloom and yield in the following year. Therefore,
flowering-site limitation, which was suggested by Goldschmidt
(2005) as one of the three major reasons for alternate bearing in
fruit trees, appears to be most relevant in olive. The developing
fruits appear to predominantly affect vegetative growth, limiting it
during the On-year. The practical consequences of the present
study are still unclear. The commercial harvest of oil olives usually
takes place not earlier than mid-October, too late for the essential
resumption of new vegetative growth. Table olives, in contrast, are
usually harvested earlier, from the beginning of September and
indeed, their tendency toward alternate bearing is generally
smaller and often correlates with time of harvest (Lavee, 1996).
Our results indicate that the opportunity to influence biennial
bearing by fruit thinning in olives is viable until at least 120 DAFB,
much later than previously estimated.
Our thanks are due to Halutza olive farm for providing their
orchard for this study.
Andreini, L., Bartolini, S., Guivarc’h, A., Chriqui, D., Vitagliano, C., 2008. Histological
and immunohistochemical studies on flower induction in the olive tree (Olea
europaea L.). Plant Biol. 10, 588–595.
Badr, S.A., Hartmann, H.T., Martin, G.C., 1970. Endogenous gibberellins and inhi-
bitors in relation to flower induction and inflorescence development in the
olive. Plant Physiol. 46, 674–679.
Baktir, I., Ulger, S., Himelrick, D.G., 2004. Relationship of seasonal changes in
endogenous plant hormones and alternate bearing of olive trees. HortScience
Connor, D.J., Fereres, E., 2005. The physiology and adaptation of yield expression in
olive. Hort. Rev. 31, 155–229.
Cuevas, J., Rallo, L., Rapoport, H.F., 1994. Crop load effects on floral quality in olive.
Sci. Hortic. 59, 123–130.
Dag, A., Bustan, A., Avni, A., Lavee, S., Riov, J., 2009. Fruit thinning using NAA shows
potential for reducing biennial bearing of ‘Barnea’ and ‘Picual’ oil olive trees.
Crop & Pasture Science 60, 1124–1130.
Fabbri, A., Benelli, C., 2000. Flower bud induction and differentiation in olive. J. Hort.
Sci. Biotechnol. 75, 131–141.
Ferna ´ndez-Escobar, R., Benlloch, M., Navarro, C., Martin, G.C., 1992. The time of
floral induction in olive. J. Am. Soc. Hort. Sci. 117, 304–307.
Goldschmidt, E.E., 2005. Regulatory aspects of alternate bearing in fruit trees. Italus
Hortus 12, 11–17 (in Italian).
Lavee, S., 1996. Biology and physiology of the olive. In: Blazquez, J.M. (Ed.), World
Olive Encyclopaedia. Plaza and Janes SA, Barcelona, Spain, pp. 61–105.
Lavee, S., 2006. Biennial bearing in olive (Olea europaea L.). Olea FAO Olive Network
Lavee, S., Harshemesh, H., Avidan, N., 1986. Phenolic acids—possible involvement in
regulatinggrowthandalternatefruitinginolive trees. ActaHortic.179,317–328.
Martin, G.C., 1989. Olive flower and fruit population dynamics. Acta Hortic. 286,
Monselise, P.S., Goldschmidt, E.E., 1982. Alternate bearing in fruittrees. Hort. Rev. 4,
Orlandi, F., Garcia-Mozo, H., Vazquez-Ezquerra, L., Romano, B., Dominguez, E.,
Galan, C., Fornaciari, M., 2004. Phenological olive chilling requirements in
Umbria (Italy) and Andalusia (Spain). Plant Biosys. 138, 111–116.
Rallo, L., Martin, G.C., 1991. The role of chilling in releasing olive floral buds from
dormancy. J. Am. Soc. Hort. Sci. 116, 1058–1062.
Rallo, L., Torreno, P., Vargas, J.A., Alvarado, J., 1994. Dormany and alternate bearing
in olive. Acta Hortic. 356, 127–136.
Ryan, D., Prenzler, P.D., Lavee, S., Antolovich, M., Robards, K., 2003. Quantitative
changes in phenolic content during physiological development of the olive
(Olea europaea) cultivar Hardy’s Mammoth. Int. J. Food Chem. 51, 2532–2538.
Sanz-Corte ´s, F., Martı ´nez-Calvo, J., Badenes, M.L., Bleiholder, H., Hack, H., Llacer, G.,
Meier, U., 2002. Phenological growth stages of olive trees (Olea europaea). Ann.
Appl. Biol. 140, 151–157.
Stutte, G.W., Martin, G.C., 1986. Effect of killing the seed on return to bloom of olive.
Sci. Hortic. 29, 107–113.
Troncoso, A., Garcia, J.L., Lavee, S., 2010. Evaluation of the present information on
the mechanisms leading to flower bud induction, evocation and differentiation.
Acta Hortic., in press.
Ulger, S., Sonmez, S., Karkacier, M., Ertoy, N., Akdesir, O., Aksu, M., 2004. Determi-
nation of endogenous hormones, sugars and mineral nutrition levels during the
induction, initiation and differentiation stage and their effects on flower
formation in olive. Plant Growth Reg. 42, 89–95.
A. Dag et al./Scientia Horticulturae 123 (2010) 469–472