Scientia Horticulturae 219 (2017) 107–117

Contents lists available at ScienceDirect

Scientia Horticulturae
journal homepage: www.elsevier.com/locate/scihorti

Synergistic effects between bumblebees and honey bees in apple

orchards increase cross pollination, seed number and fruit size
G. Sapir a,b , Z. Baras a,b , G. Azmon a,b , M. Goldway a,b , S. Shafir c , A. Allouche d , E. Stern d ,
R.A. Stern a,b,∗
a
MIGAL, Galilee Research Institute, P.O. Box 831, Kiryat Shmona 11016, Israel
b
Department of Biotechnology, Faculty of Life Sciences, Tel-Hai College, Upper Galilee, 12210, Israel
c
Department of Entomology, Faculty of Agriculture, Food and Environment, Hebrew University of Jerusalem, Rehovot, Israel
d
BioBee Ltd., Sde Eliyahu, Israel

a r t i c l e i n f o a b s t r a c t

Article history: Most apple cultivars are self-sterile and completely dependent on cross-pollination from a different
Received 8 December 2016 cultivar in order to set fruit. Various insects may be pollinators, but the main one is the honey bee [HB]
Received in revised form 1 March 2017 (Apis mellifera). However, despite the advantages of the honey bee as pollinator of many plants, it is a
Accepted 2 March 2017
relatively inefficient pollinator of apple flowers. The main reason for this is the tendency of HBs to visit the
Available online 11 March 2017
apple flower from the side (sideworker), thus “stealing” nectar without touching the flower’s reproductive
organs – stamens and stigma. In contrast, a bee that visits the flower from the top (topworker) contacts
Keywords:
the flower’s reproductive organs, which results in better pollination. Due to the low pollination efficiency,
Honey bee
Bumblebee
few seeds are formed, and often the resulting fruit is too small to be of commercial value. Experiments
Apple (Malus domestica) conducted in Israel over the last few years have shown for the first time that adding bumblebees [BB]
Insect foraging behavior (Bombus terrestris) into pear orchards improved cross-pollination, thus increasing the number of seeds
and subsequently fruit size. The goal of the present work was to test the hypothesis that adding BBs
to apple orchards may improve cross-pollination. We found that adding BBs to the HBs in the apple
orchard improved pollination in all tested cultivars, especially in ‘Gala’, which naturally suffers from
relatively few seeds in the fruit. It appears that the addition of BBs did not only increase the number of
pollinating insects in the orchard that could perform cross-pollination, including in the cool mornings and
in adverse weather conditions, but that it also changed HB foraging behavior, which resulted in improved
cross-pollination and increased efficiency, and subsequently more seeds and larger fruit. The improved
pollination was due to the greater mobility of HBs between rows of pollinated cultivar and pollenizer,
and to the greater proportion of topworkers, which are more efficient pollinators.
© 2017 Elsevier B.V. All rights reserved.

1. Introduction most common yield-limiting factors (Free, 1993; Goulson, 2010;

Most apple cultivars are self-incompatible (Dennis, 2003;
Goldway et al., 2007; Schneider et al., 2001, 2005). Hence, for effec-
tive pollination and successful fertilization they depend entirely on
the synchronization of flowering among different cultivars of trees
and on the intensive activity of pollinators. The most important pol-
linators of apple are honey bees (HB; Apis mellifera) (Delaplane and
Mayer, 2000; Dennis, 1979; Free, 1993). Since apple trees are usu-
ally grown in temperate zones where weather conditions during
the blooming period may be unfavorable for bee flight, insufficient
pollination, pollen tube growth, and fertilization are among the
∗ Corresponding author at: MIGAL, Galilee Research Institute, P.O. Box 831, Kiryat
Shmona 11016, Israel.
E-mail address: raffi@migal.org.il (R.A. Stern).
Hoopingarner and Waller, 1993).
Honey bees often switch from apple flowers to the compet-
ing flowers, which they find more attractive and rewarding (Free,
1993). Moreover, HBs tend to restrict their mobility to one row of
trees, which usually contains a single cultivar (Stern et al., 2001). In
addition, the Effective Pollination Period (EPP; i.e., ovule longevity,
or the time between pollination and fertilization) is very short in
apple, lasting for only 1–2 days (Dennis, 1979). Thus, although the
stigma remains receptive for longer periods, pollination needs to
be accomplished within 1–2 days since onset of anthesis for fertil-
ization to occur before degeneration of the ovule.
Another problem with apple pollination is that HB activity on
apple flowers is not always efficient. They collect both nectar and
pollen from the flower, but not necessarily at the same time (Mayer,
1984). Usually, when collecting pollen, they pollinate the flower

http://dx.doi.org/10.1016/j.scienta.2017.03.010
0304-4238/© 2017 Elsevier B.V. All rights reserved.
108 G. Sapir et al. / Scientia Horticulturae 219 (2017) 107–117
effectively, because they work from the top of the flower as “top-
workers”, but they often do not collect nectar from the top. There
are gaps at the base of the stamens, especially on ‘Red Delicious’,
that enable “sideworking” to obtain nectar without contacting the
anthers and stigmas of the flower (Dennis, 1979; Roberts, 1945;
Robinson, 1979). Sideworking is efficient for the honey bee, but
requires time to learn (DeGrandi-Hoffman et al., 1985).
Due to these various factors, pollination of apple is especially
inefficient, especially in ‘Red Delicious’ (Stern et al., 2001; Schneider
et al., 2002). Any technique that can improve HB mobility between
rows for better cross pollination and increase HB efficiency by
working as topworkers would improve apple yield and especially
fruit size (through more seeds in the fruit).
In previous work (Stern et al., 2001) it was shown that
the sequential introduction of HB colonies into apple orchards
increases bee activity on trees, bee mobility between rows, rate
of topworkers, and as a result improves cross pollination, fruit size
and yield. However, the fruit size remained small due to low (under-
optimal) seed number (7–9 seeds fruit−1 ).
Evidence has been accumulating in recent years showing a pos-
itive relationship between diversity and ecosystem services, like
crop pollination (Klein et al., 2007, 2012). For some crops, wild bees
are more effective pollinators on a per visit basis than HBs (Willmer
et al., 1994; Zhang et al., 2015) and can functionally complement
the dominant visitor (Albrecht et al., 2012). Interactions between
floral visitors may modify their behavior through interference com-
petition (Greenleaf and Kremen, 2006) or resource competition
(Inouye, 1978). Either form of competition may augment pollina-
tion. For example, due to interference competition, interaction with
non-Apis bees caused HBs to move more often between rows of
sunflower (Helianthus annuus), thereby increasing their pollination
efficiency (number of seeds production per visit) (Greenleaf and
Kremen, 2006). Resource competition can also alter pollinator for-
aging movement (Inouye, 1978). Another form by which foraging
behavior may be differentially affected by conspecific or interspe-
cific interactions is through flower marking. For example, HBs are
more likely to avoid visiting a flower recently visited by a bum-
blebee than by another HB (Stout and Goulson, 2001). Changes in
pollinator movement are particularly important in crop species that
are self-incompatible like apple and pear.
The buff-tailed bumblebee (BB; Bombus terrestris L.) is another
potential vector for pollination in apple (Goulson, 2010), although
it is most commonly used in vegetable greenhouses to pollinate
tomato, strawberry, and pepper (Dimou et al., 2008; Kearns and
Inouye, 1997; Pressman et al., 1999; Van den Eijnde et al., 1991;
Velthuis and Van Doorn, 2006). BB possess several attributes that
may be effective for outdoor pollination of field crops and orchards.
As with HBs, BBs collect nectar and pollen from numerous sources
(Heinrich, 2004). However, their capacity to carry nectar and pollen
is greater than that of HBs (Free and Williams, 1972) as their body
is about twice the size of the HB body. By carrying greater loads per
foraging bout, a BB would make more floral visits, thus increasing
the probability of switching between rows within a bout, which
could make it a more efficient pollinator (Heinrich, 2004). BBs
are very active, generally visiting many more flowers than HBs
(Goodell and Thomson, 1997; Willmer et al., 1994). For example,
the level of cross-pollen in ‘Conference’ pear trees pollinated by
BBs was twice that in their HB-pollinated counterparts (Jacquemart
et al., 2006). We found in apples that BBs visited about 40 flow-
ers min−1 , whereas HBs visited only about ten flowers min−1 (R.A.
Stern, unpublished data). In addition, the range of activity of BBs is
usually restricted to a few hundred m from the hive, increasing the
chance of pollination within the orchard, whereas HBs are active
over thousands of m from the hive (Wolf and Moritz, 2008).
BBs are active at temperatures below 14 ◦ C, which is the limiting
temperature for HB activity (Vicens and Bosch, 2000). Thus, BBs can
commence foraging earlier in the day than HBs. Furthermore, in
contrast to HBs, BBs can forage under harsh winter condition. They
have even been observed to forage in wind and rain (Corbet et al.,
1993; Goulson, 2010; Lundberg, 1980; Tuell and Isaacs, 2010).
BBs do not signal the location of floral resources as HBs do,
and therefore, unlike HBs, they do not recruit nest mates out of
the orchard to competing wild flowers in surrounding meadows
(Heinrich, 2004).
The use of BBs as commercial pollinators has mainly been
applied in greenhouses and, to a minor extent, in open fields
of alfalfa (Medicago sativa) and red clover (Trifolium pretense).
Recently we reported the first commercial use of BBs in pear
orchards (Zisovich et al., 2012). Observations of BBs in apple
orchards have been reported (Goodell and Tomson, 1997), but not
their commercial application. The aim of the present study was
to evaluate the effects of adding hives of Bombus terrestris to HB
colonies in apple orchards on cross-pollination, seed number per
fruit and fruit size.
2. Materials and methods
2.1. Orchard design
The experiments were conducted in four commercial apple
(Malus domestica) orchards located in the Upper Galilee (Baram
– 700 m asl) and in the Golan Heights (Ortal, Elrom and Ramat
Magshimim – 1000 m asl) in the north part of Israel (the main
experiments were done in Baram). The cultivars tested were ‘Gala’,
‘Red Delicious’, ‘Golden Delicious’ and ‘Pink lady’ grown on MM
106 rootstocks. The trees in each orchard were uniform in age (ca.
10 years) and size, and had similar crop loads in the year prior
to the experiments. Two rows of the same cultivar were planted
at spacing of 2.5 m between trees and 4.5 m between rows (900
trees ha−1 ) with an adjacent 2 rows of another cultivar on each
side as a pollenizer, throughout the entire orchard. The rows were
always aligned in the North-South direction. There was a good over-
lap between the flowering times of the cultivars in each year. As
in previous experiments with HBs (Stern et al., 2001, 2004) and
BBs (Zisovich et al., 2012), the treatments were kept in the same
orchard, which was divided into two areas: HB + BB [+BB] vs. HB
only [−BB] = control, with a distance of ca. 1000 m between the
two areas. Thus, treatment effects could be compared while main-
taining other parameters such as cultivars, rootstocks, agricultural
practices, climate conditions, competing flora, etc., as similar as
possible.
2.2. Bees treatments
HB and BB colonies were distributed homogenously through-
out the orchard. HB colonies (Kibbutz Dan Apiary, Upper Galilee,
Israel) were introduced into each orchard according to commercial
recommendations developed in Israel, following Stern et al. (2001).
The first introduction was at 2.5 colonies ha−1 at 10% full bloom (FB)
and the second introduction was an additional 2.5 colonies ha−1 at
FB, for a final total of 5.0 HB colonies ha−1 . A HB colony reaches up
to 30,000–40,000 bees, whereas a BB colony reaches only 150–300
bees.
The BB hives (BioBee Ltd., Kibbutz Sde-Eliyahu, Emek-
Hamaayanot, Israel) were introduced into the orchard at a density
of ten hives ha−1 at the start of bloom (i.e. 1–2 days before the intro-
duction of HB). From our experience, it takes BB colonies a few days
to acclimate to the orchard and to start foraging on the trees. Since
BBs are less flower constant (Goulson, 2010), they are probably able
to strengthen the colony by foraging on competing plants outside
 G. Sapir et al. / Scientia Horticulturae 219 (2017) 107–117
the orchards for a few days, and then shift to the apple when it
starts to bloom.
2.3. HB and BB foraging activity
HB and BB foraging activity on apple trees was assessed at ca.
50 m from each hive by an observer who counted the bees on a
whole tree, using a hand-held counter, while circling the tree at a
distance of ca. 1.0 m for 60 s (Stern et al., 2001, 2004).
The number of HBs or BBs per tree was measured almost every
morning (08.00–10.00 h) on ten trees per treatment in each exper-
iment for each observation.
Bee mobility between adjacent rows (two different cultivars)
was assessed in the same trees by an observer with their back
to the sun, who used a manual counter to count all the bees that
crossed between the observation trees at observing distance. Each
observation lasted 60 s, and ten observations during the morning
were made on each observation date, in each experiment. In most
replicates (Baram, 2013, Ortal, 2013 and Baram, 2014) we did not
distinguish between HB and BB when counting bee mobility. In a
couple of replicates (Baram and Ortal, 2013) we conducted an addi-
tional assessment of only HB mobility. The method was similar to
that above, but to be able to focus on HBs, we monitored mobility
between a single tree in one row to a single tree in another row.
2.4. Foraging behavior
Foraging behavior was categorized as sideworking or topwork-
ing. Sideworking was defined when a bee had all of its legs on the
petals, and it imbibed nectar by forcing its tongue between the base
of the stamens toward the nectary. During topworking, the bees’
legs were on the stamens while it collected nectar or pollen. Nec-
tar was imbibed by the bee probing down its tongue through the
anthers to the nectary. A bee that collected pollen would stay on
top of the flower, collecting pollen from the anther with its legs
(Kuhn and Ambrose, 1982; Stern et al., 2001). In each treatment,
in each orchard, on each cultivar, on each date, we monitored the
foraging activity of fifty haphazardly selected HBs foraging on the
apple trees, and categorized them into sideworker or topworker,
collecting nectar or pollen.
2.5. Seed number per fruit and fruit size
In each replicate of the experiment, we selected ten trees in each
treatment. The trees
had uniform flowering intensity, and were adjacent to rows with
the optimal pollenizer (full genetic compatibility with the target
cultivar – Goldway et al., 2007). On each of these Trees 20 king
flowers (200 flowers per treatment) were labeled at FB. At harvest,
the remaining labeled fruit (100–130) were picked and fruit size
and seed number per fruit (in the same fruit) were recorded.
Statistical analysis
Percentage data were subjected to arcsin transformation before
analysis, to provide a normal distribution. Data were analysed for
statistical significance by the general linear model (GLM) proce-
dure. Duncan’s new multiple range test was applied to compare
treatments when ANOVA showed significant differences among the
means at P ≤ 0.05. For correlations, we report the Pearson product-
moment correlation coefficient, r.
109
3. Results
3.1. Bee activity on trees
Adding BBs to the apple orchards usually had no influence on HB
activity during 2012 and 2013 at Baram (Fig. 1A and B). The number
of HBs tree−1 min−1 was similar for the control (−BB) and BB treat-
ment (+BB) on each measurement day during bloom. It reached a
maximum of 12 HB tree−1 min−1 at FB in both treatments, which
is the usual activity on apple trees at FB (Stern et al., 2001). Dur-
ing most of the bloom period the temperature was optimal for HB
activity (daily mean maximum of 25 ◦ C). However, on cool days
in which the maximum temperature was below 15 ◦ C, i.e. 11 April
2012 (Fig. 1A) and 4, 9 and 11 April 2013 (Fig. 1B), HB activity was
very low (only ca. 4 bees tree−1 min−1 in both treatments) while BB
activity was not reduced on the same days (Fig. 1D and E). However,
in 2014 HB activity on the trees was lower in the BB treatment com-
pared to the control (Fig. 1C). In all years, in all the other orchards
(Elrom, Ortal, Ramat Magshimim), the pattern of HB activity (not
shown) was similar to that of Baram 2012 and 2013 (Fig. 1A and B).
In contrast to HB activity in both treatments, BB activity was
much higher in the +BB treatment than in the control (−BB) in all
three years (Fig. 1D–F). Similar results were obtained in all other
experiments in all orchards and years (data not shown).
3.2. Bees mobility between rows
In preliminary experiments that we conducted between 2009
and 2012, we monitored the number of bees per tree, as we had
previously done in pear (Zisovich et al., 2012). With these data,
we were unable to explain the increase in the +BB treatment in
fruit size and yield, which indicate improved fertilization. There-
fore, since 2013 we began monitoring also bee mobility, the number
of bees crossing between two adjacent rows of different cultivars, as
an assessment of cross-pollination potential. At first, we recorded
the total number of bees crossing per minute observation, without
discriminating between HBs and BBs. In the main experiment in
Baram, 2013, we found during the entire blooming period, and in
each of the days of observations, that bee mobility (HBs and BBs)
was always greater in the bumblebee treatment (+BB) than in the
control (−BB). The number of bees crossing was often more than
double in the +BB treatment compared to the control, with between
4 and 10 in comparison to only 2 (expect on April 8), respectively
(Fig. 2A). We found a similar pattern in 2014 in the same orchard,
with about double the bee mobility in the +BB treatment than in
the control (Fig. 2B). Similar results were obtained also in the Ortal
orchard, 2013 (Fig. 2C), and in other orchards in which bee mobility
was assessed for only four days, such as Ramat Magshimim, 2013,
Elrom, 2014 and Baram, 2015 (data not shown).
In order to assess whether the greater bee mobility in the +BB
treatment was due only to the addition of bees (BBs) to the orchard
or possibly also due to increased mobility of HBs, in 2013 we con-
ducted more detailed observations (see Section 2) in which we
discriminated between HB and BB crossings. These detailed obser-
vations were conducted in only two orchards, over two days during
full bloom, but they clearly show a significant, two-fold or more,
increase in HB mobility in the +BB treatment (Fig. 3). Based on
these findings, we assume that the total increased mobility in the
+BB treatment in our other experiments (Fig. 2), is largely due to
increased mobility of HBs, and not just due to mobility of BBs.
3.3. Honey bee flower handling behavior
Another means by which apple cross-pollination by honey bees
can be increased is by increasing the proportion of HB foragers that
collect pollen or nectar as topworkers, rather than as sideworkers,
110 G. Sapir et al. / Scientia Horticulturae 219 (2017) 107–117

Fig. 1. Effect of adding bumblebee (BB) hives to honey bee (HB) colonies [+BB treatment] on the average number of HB (Panels A–C) and BB (Panels D and E) visits tree−1 min−1 .
Data of each column are the means ± S.E. of ten replicate trees per treatment (“ + BB” vs. “−BB” = control) at the Baram orchard in 2012–2014. FB, Full bloom was on 9/4/2012,
7/4/2013 and 6/4/2014. *Significant at P ≤ 0.05. X = Maximum temperature below 15 ◦ C.

which are inefficient pollinators (Stern et al., 2001). Consequently,
in 2013, we assessed the proportion of topworkers in three differ-
ent orchards: Baram, Ortal, and Ramat Magshimim (Fig. 4). In all
three orchards, and over most of the blooming period, proportion
of topworkers was higher in the +BB treatment than in the con-
trol. Pollen collectors are always topworkers, but nectar collectors
can be either topworkers or sideworkers. In order to further eval-
uate whether the increase in topworkers was due to more pollen
collectors, or due to an increased proportion of topworking nectar
collectors, in 2014 we monitored HB foraging behavior in the Baram
orchard. Among the pollen collectors, the proportion of topwork-
ers was similar in the +BB treatment and in the control, but among
the nectar collectors, the proportion of topworkers was about 50%
greater in the +BB treatment relative to the control (Table 1).
3.4. Seed number and fruit size
Increasing cross-pollination in apple orchards (Figs. 1–4 and
Table 1) should improve the fertilization process and thereby
increase the number of seeds per fruit. We therefore counted seed
numbers in 2014 in fruit of two cultivars in each of three orchards
(Table 2). Mean number of seeds per fruit was greater in the +BB
 G. Sapir et al. / Scientia Horticulturae 219 (2017) 107–117 111

Fig. 2. Effect of adding bumblebee (BB) hives to honeybee (HB) colonies [+BB treatment] on the average number of bee mobility (HB + BB together) between rows min−1 .
Data of each column are the means ± S.E. of ten observations per treatment min−1 (“+BB” vs. “−BB” = control) at Baram 2013 (A), Baram 2014 (B) and Ortal 2013 (C). FB, Full
bloom was on 7/4/2013, and 6/4/2014 at Baram and 10/4/2013 at Ortal. *Significant at P ≤ 0.05.

Table 1 statistically significant. We attribute this to the fact that (by mis-
Effect of adding bumblebee (BB) hives to honey bee (HB) colonies [+BB treatment] on
take) the rows of Gala in the +BB treatment, which were marked
the percentage of HBs that collected nectar from the side (sideworking) or from the
top (topworking), or that collected pollen (which is always by topworking). Data are and from which fruit was analyzed, were adjacent to rows of
mean percentages for 50 HBs (10 HBs per replicate × 5 replicates) in each treatment Golden Delicious. This cultivar is genetically only semi-compatible
in each of 3 days during FB at Baram orchard, 2014. with Gala (Schneider et al., 2005). In contrast, the rows of Gala
Treatment HB foraging activity (%) Total
in the control (-BB) plots were adjacent to rows of Red Delicious,
which is genetically fully-compatible with Gala, and therefore with
Pollen collection Nectar collection
greater fertilization potential as a pollinizer (Schneider et al., 2005;
Topworking Sideworking Topworking Goldway et al., 2007). Interestingly, there were not fewer seeds per
+BB 10 a 32 b 58 a 100 fruit in the +BB plot, despite the only semi-compatible pollinizer
−BB 11 a 49 a 40 b 100 in this plot. It therefore appears that the loss in fertilization due
Mean values within a column followed by different letters differ significantly by
to the semi-compatibility of the pollinizer can be compensated by
Duncan’s new multiple range test, P ≤ 0.05. the gain in fertilization due to addition of BBs. Also apparent from
Table 2, is that the lower the number of seeds in the −BB control,
the greater the increase in number of seeds with the addition of
treatment than in the control in both cultivars in all three orchards. BBs (not including the Gala plot in Baram: r = 0.98, P = 0.004).
Only in the Gala cultivar in Baram was the difference small and not
112 G. Sapir et al. / Scientia Horticulturae 219 (2017) 107–117

Fig. 3. Effect of adding bumblebee (BB) hives to honeybee (HB) colonies [+BB treatment] on the average number of HB mobility between rows min−1 . Data of each column
are the means ± S.E. of 20 observations per treatment min−1 (“+BB” vs. “−BB” = control) at the Baram (A) and Ortal (B) orchards in 2013. FB, Full bloom was on 7/4/2013 at
Baram and 10/4/2013 at Ortal. *Significant at P ≤ 0.05.

Table 2 Table 3
Effect of adding bumblebee (BB) hives to honey bee (HB) colonies [+BB treatment] Effect of adding bumblebee (BB) hives to honey bee (HB) colonies [+BB treatment]
on the number of seeds per fruit and fruit size. The experiment was conducted in on the number of seeds per fruit in ‘Gala’ according to the pollinizer. The experiment
2014, with two cultivars in each of three orchards. was conducted in 2014 at Ortal orchard.

Orchard site Cultivar Seed no.1 Fruit size (mm)1 Bee treatment Pollenizer Compatibility Seed no./fruit1

+BB −BB +BB −BB +BB Red Delicious Full 7.5 a

Golden Delicious Semi 7.0 a
Baram Gala 6.3 a 6.2 a 68 a 68 a
Golden Delicious 9.1 a 8.4 b 70 a 69 a −BB Red Delicious Full 6.1 b
Granny Smith Full 5.9 b
Elrom Gala 7.2 a 4.0 b 72 a 69 b
Pink Lady 8.2 a 7.6 b 74 a 71 b Mean values within a column followed by different letters differ significantly by
Duncan’s new multiple range test, P ≤ 0.05.
Ortal Gala 7.9 a 6.1 b 73 a 69 b 1
Data are the mean of 100 fruit (10 fruit per tree × 10 trees with the same crop
Red Delicious 8.1 a 6.3 b 74 a 72 b
load).
Mean values within a row, for each parameter, followed by different letters differ
significantly by Duncan’s new multiple range test, P ≤ 0.05.
1
Data are the mean of 100 fruit (10 fruit per tree × 10 trees with the same crop size (Table 2). There was a statistically significant (P < 0.05) cor-
load) selected randomly from each cultivar in each orchard (Baram, Elrom, Ortal) relation between seed number and fruit diameter in each of the
and in each treatment (+BB vs. −BB).
two cultivars in all three orchards (Fig. 5). Each additional seed
increased fruit diameter by between about 0.5 and 1.2 mm. The
Further support for the more significant effect of the +BB greatest effect was in Gala in the Ortal orchard, where each addi-
treatment relative to that of a fully-compatible pollinizer on cross- tional seed increased fruit diameter by 1.2 mm (y = 1.2x + 61.9). For
pollination and seed number was obtained from an additional example, an apple with 3 seeds reached a diameter of 65 mm, which
experiment that we conducted on Gala in the Ortal orchard with is below economic threshold for this crop, whereas with 7 seeds the
different pollinizers (Table 3). When the pollinizer was the fully- diameter was 70 mm (1.2 mm x 4 seeds = 5 mm), which is above
compatible Red Delicious, there were more seeds in Gala apples economic threshold and yields a high price. In Gala apples, an eco-
in the +BB treatment, showing again the effect of adding BBs. nomically valuable fruit of 70 mm is reached when there are 7 or
However, seed number was greater even in the +BB plot with the more seeds (Fig. 5, Gala in all three orchards). Therefore, growers
semi-compatible Golden Delicious as pollinizer relative to the −BB aim at getting a high proportion of fruits with at least 7 seeds.
plots with either of two fully-compatible pollinizers. From the distribution of fruit with different number of seeds
The greater mean number of seeds per fruit in the +BB treat- (Fig. 6), it can be seen that most of the fruit with few (1–5) seeds
ment was generally reflected (except in Baram) also in greater fruit were in the control (-BB) plots, whereas most of the fruit with many
 G. Sapir et al. / Scientia Horticulturae 219 (2017) 107–117
Fig. 4. Effect of adding bumblebee (BB) hives to honeybee (HB) colonies [+BB treatment] on the proportion of “Topworkers”. Data of each column are the means ± S.E. of 50
selected HB in each treatment in each date in each orchard. Baram (A), Ortal (B) and Ramat Magshimim (C), 2013. *Significant at P ≤ 0.05.
(7–11) seeds were in the bumblebee treatment (+BB). Fruit with 7 or
more seeds was obtained in 62% of the apples in the +BB treatment,
but in only 35% (just a bit over one half as many) of the apples in
the −BB control. Similar distributions of fruit with more seeds in
the +BB treatment, were obtained in the same year (2014) also in
Golden Delicious in Baram, in Red Delicious in Ortal, and in Pink
Lady in Elrom (Fig. 7).
4. Discussion
Small fruit size, even after reducing crop load by thinning (Stern,
2015), is usually the result of low seed number in the fruit (Stern
et al., 2001; Tromp and Wertheim, 2005; Westwood, 1993). Fruit
with few seeds tends to drop, and those that remain attached to
the tree are inclined to be small due to impaired development. This
phenomenon has been well documented in many fruits, among
them Asian pear (Pyrus pyrifolia), European pear (Pyrus communis),
strawberry (Fragaria x ananassa), mandarin (Citrus) and kiwifruit
(Actinidia deliciosa) (Dimou et al., 2008; Goulson, 2010; Schneider
113
et al., 2009; Zhang et al., 2010; Zisovich et al., 2010). The results from
5 years of preliminary observations (2007–2011; data not shown)
and from the current experiments, especially in 2013 and 2014
(Table 2 and 3 and Figs. 5 and 6), indicate that the average number of
seeds in each apple cultivar, and especially in ‘Gala’ fruit, is under
the optimum of ten seeds per fruit known to support rapid and
optimal fruit development (Tromp and Wertheim, 2005) and also
good fruit storage (Westwood, 1993). In previous reports on Euro-
pean pear we showed that an increase in the number of seeds can
be achieved by pollination with fully compatible cultivars and by
improving HB activity (Stern et al., 2004, 2007; Zisovich et al., 2004,
2005). We received similar results with fully compatible cultivars
in apple (Schneider et al., 2001, 2005; Stern et al., 2001).
We also found that adding BBs to the orchard, in addition to HBs,
increased the number of seeds per fruit in Spadona pear from two
or three to between four and six seeds (Zisovich et al., 2012). In the
present study, we tested whether a similar phenomenon occurs
in apple. We introduced HB colonies sequentially up to a stan-
dard application of 5.0 hives ha−1 (since a higher density did not
114 G. Sapir et al. / Scientia Horticulturae 219 (2017) 107–117

Fig. 5. Relationship between the number of seeds per fruit and the average diameter of the fruit. Data are based on the data presented in Table 2 from both treatments (+BB
and −BB together). Each data point is an average of all fruit with the indicated number of seeds.

Fig. 6. Distribution of the percentage of ‘Gala’ fruit with 0–11 seeds per fruit in the two treatments (+BB and −BB). The data of each column are the means ± S.E. of 3 orchards
(Baram, Ortal and Elrom) in 2014 (and based on the data presented in Table 2).

give better results) (Stern et al., 2001), and examined the impact of
adding BBs. We hypothesized that the addition of BBs would con-
fer an advantage for cross-pollination and increase seed number
per fruit and fruit size, especially under sub-optimal conditions for
pollination, including temperatures below 15 ◦ C, wind, clouds, rain,
non-synchronous blooming, and competition by wild flowers. Such
conditions are typical for several crops in the Rosaceae. Improved
pollination due to the addition of other bee species to pollination
by HBs has been reported also for almond (Brittain et al., 2013), and
even peach, which is self-compatible (Zhang et al., 2015). Interest-
ingly, in these latter fruits there is only one seed, but it is crucial for
fruit development.
As hypothesized, in most of the experiments in which BBs were
added, the number of seeds per fruit increased. The seed number in
 G. Sapir et al. / Scientia Horticulturae 219 (2017) 107–117
Fig. 7. Distribution of the percentage of ‘Golden Delicious’ (Baram), ‘Red Delicious’ (Ortal) and ‘Pink Lady’ (Elrom) fruit with 0–11/0–14 seeds per fruit in the two treatments
(+BB and −BB) during 2014.
apple fruit is higher than in pear, and there were many fruits with 6
or more seeds even in the −BB treatment = control (Tables 2 and 3
and Fig. 6). The lower the number of seeds per fruit in the control,
the greater was the contribution of adding BBs to increasing seed
number. For example, the Gala cultivar suffered from the lowest
number of seeds (DeGrandi-Hoffman et al., 1985) in the control, and
the number increased by a mean of 3.2 seeds (80%) in the +BB treat-
ment (Table 2). We generally found increases in number of seeds
per fruit in the +BB treatment in all the preliminary experiments
we conducted with different cultivars in the years 2009–2011.
Bumblebees are well adapted to cooler environments and
can forage efficiently also at ambient temperatures below 15 ◦ C
(Goulson, 2010). Such cool conditions occur often in Israel during
the apple bloom. In contrast, HBs hardly forage at these ambi-
ent temperatures (Zisovich et al., 2012; Vicens and Bosh, 2006).
For example, on April 11, 2012 (Fig. 1A) and on April 4, 9, and
11, 2013 (Fig. 1B), HB activity was low (2–4 bees/tree/min) when
ambient temperatures were below 15 ◦ C, in comparison to 10–12
bees/tree/min on warmer days. BBs on the other hand, were not
affected by the cool weather during these days (Fig. 1D and E).
115
We noted constant BB activity of about 1–2 bees/tree/min, at
different times from 0600 to 1200. We similarly measured high
BB activity at the hive, of about 8–10 bee exits and entrances per
min, both during the cooler early morning (0600–0800) and during
the warmer later hours (0800–1200). We do not show these data
here, but they are similar to those found in our experiments in pear
orchards (Zisovich et al., 2012). HBs, in contrast, usually began for-
aging about two hours after the BBs, at around 0800, when ambient
temperature reached 15 ◦ C.
Increasing cross-pollination results in greater fertilization suc-
cess and greater number of seeds per fruit. We believe that rates of
cross-pollination increased in the +BB treatment due to two main
factors: the introduction into the orchard of an additional (vector)
pollinator that is active at lower ambient temperatures, and the
interaction between the two bee species. Consistent with previous
reports (Vicens and Bosch, 2000), we found that BBs were active in
the early cool morning hours. The interaction between BBs and HBs
resulted in greater mobility of HBs between rows and in more top-
working, which is more efficient for pollination. The effect of BBs on
HB behavior could be direct, through interference competition, in
116 G. Sapir et al. / Scientia Horticulturae 219 (2017) 107–117
which two bees meet at a flower and the BB physically disrupts the
foraging of a HB, making her shift to a different row (Greenleaf and
Kremen, 2006). Or it could be indirect, through resource competi-
tion, in which both bee species consume the same nectar and pollen
resources (Brittain et al., 2013). We made about 1000 observations,
during each of the days with bloom during the 2013 season, of HBs
foraging on flowers to assess the degree of disruption by BBs. We
noted only one encounter between a HB and a BB (and even then,
it did not make the HB shift away). It therefore appears that the
greater mobility of HBs is due to resource competition.
Due to time constraints, we did not compare pollen and nec-
tar amounts in flowers in the +BB and −BB orchards. However, in
line with the resource competition hypothesis, we believe that the
early foraging of BBs on the apple flowers would reduce the nectar
standing crop and/or the available fresh pollen that remains in the
orchard. The number of BB foragers in the orchard is lower than
that of the HBs, but they visit flowers at a high rate, and their crop
capacity is about four times greater than that of HBs, so in two
hours they could make a significant impact on available resources.
When the HBs would arrive at these flowers, they would find a more
variable distribution of pollen and nectar, with some flowers con-
taining less food, if any. HBs dislike variable resource distributions,
especially when they also contain empty flowers (Drezner-Levy and
Shafir, 2007). While generally preferring to visit flowers that are
nearby and in a straight row, after visiting empty flowers, HBs tend
to move further and more to the side (Waddington, 1980). Thus,
they would be more likely to switch to another row, containing a
different cultivar.
We in fact found greater (about double) bee mobility between
rows in the +BB treatment than in the −BB control. This effect
repeated itself in all of our experiments (including those for which
data are not presented here), in all orchards, years and days of
bloom (Fig. 2A–C). However, in these observations we counted HBs
and BBs together, and we could not discriminate between mobility
of the two bee species. The effect of greater bee mobility on increas-
ing cross-pollination was clear, but these data could not test our
hypothesis that the foraging behavior of the HBs changed in the +BB
treated orchards. In an additional experiment in 2013, we therefore
counted specifically the mobility of HBs between rows. We found
that HB mobility was in fact about double in the +BB treatment
relative to the −BB control (Fig. 3). Brittain et al. (2013) recently
reported a similar finding in almond orchards, where HB mobility
between rows of different cultivars was greater when other species
of wild bees (including BBs and Osmia) were present in the orchard
than when they were absent. These authors further showed that
despite almond having only one seed, with increased bee mobil-
ity and cross-pollination, more foreign pollen (of the pollinizer
cultivar) reached the stigma, germinated, and grew pollen tubes.
Consequently, fruit-set was greater in the orchards in which there
were other bee species in addition to HBs. Fruit size was not mea-
sured in that study.
Another effect of the +BB treatment on HB behavior was an
increase in the proportion of HBs that visited flowers from the
top, as topworkers. We found this effect in all three orchards in
which this was tested in 2013 (Fig. 4). In an additional experiment
in Baram, 2014, we also noted whether bees were collecting nectar
or pollen. Only about 10% of flower visits were for pollen collec-
tion and this percentage was not affected by the +BB treatment
(Table 1). However, among the nectar collectors, the proportion
of topworkers increased by about 50% in the +BB treatment rela-
tive to the −BB control (Table 1). A possible explanation for this
effect is that nectar-collecting HBs engage in sideworking in order
to avoid the disruption of being dirtied by pollen when topworking.
In the +BB treatment, BBs would collect much of the pollen from
the flowers early in the morning. HBs would then be more likely
to engage in topworking from flowers that release less pollen, and
would be able to reach the nectaries with their proboscis with-
out being excessively dirtied by pollen. The two effects of the +BB
treatment were increasing bee mobility and increasing the propor-
tion of topworkers. The former clearly improves cross-pollination.
Topworking is generally also associated with greater pollination
effectiveness (Stern et al., 2001). However, the effectiveness of
floral visits, including topworking, would decline if flowers were
previously greatly cleaned of pollen by BBs. Such is sometimes
the case in passionfruit, for example (Ish-Am, 2009). We do not
know the relative contribution to improved cross-pollination in
our experiments of the two effects. In our experiments, BBs prob-
ably did not completely deplete a significant proportion of flowers
from pollen, since overall their presence resulted in greater cross-
pollination. The +BB treatment resulted in improved fertilization
and subsequently more seeds per fruit in all three orchards and all
cultivars (Tables 2 and 3, and Fig. 6). The only exception was the
Gala cultivar in Baram, in which by accident in the +BB treatment
we chose a row with an only semi-compatible pollinizer, whereas
in the −BB control there was a fully-compatible pollinizer. But even
there, we did not observe a decline in seed number.
The experiment in the Ortal orchard in 2014 with the Gala cul-
tivar further showed the dramatic effect of the +BB treatment in
improving pollination and increasing seed number, which was even
greater than the effect of having a fully-compatible cultivar relative
to a semi-compatible one. The semi-compatible cultivar, Golden
Delicious, nevertheless contributed to greater seed number in Gala,
when it was in an orchard with BBs, relative to the contribution
of the fully-compatible cultivar, Granny Smith, but in an orchard
without BBs (Table 3).
The direct benefit of greater seed number was increased fruit
size (Table 2). The significant and positive correlation between seed
number and fruit size in the various experiments, showed that each
additional seed contributed on average about 1 mm to fruit diame-
ter (Fig. 5). For example, in the Gala cultivar in Ortal, an apple with
three seeds yielded fruit diameter of 65 mm, which is below the
economic threshold, whereas with seven seeds fruit size reached
a diameter of 70 mm, which is around the threshold of economic
profitability. Of course, ten seeds already yielded a very large fruit,
with diameter of 74 mm.
In all cultivars studied, the distribution of fruit was skewed
towards those with a greater number of seeds in the +BB treat-
ment and towards those with a lesser number of seeds in the −BB
control (Figs. 6 and 7). In Gala for example, for between 1 and 5
seeds per fruit, there were more such fruits in the control than in
the +BB treatment, but for between 7 and 11 seeds per fruit, there
were more in the +BB treatment than in the control (Fig. 6). The
proportion of fruit with 7–11 seeds was almost double in the plots
with HBs and BBs relative to those with only HBs (64% and 35%,
respectively).
5. Conclusion
The addition of BBs as a vector for pollination was shown to
be effective in all apple cultivars, and especially in a cultivar like
‘Gala’, which suffers from insufficient number of seeds per fruit. The
addition of BBs increased the number of pollination vectors in the
orchard, and added pollination activity of bees in the cold morning
hours and/or on days with winter conditions. Most interestingly,
the BBs also changed the behavior of the HBs, thereby increas-
ing their pollination efficiency. This was due to the HBs increasing
their mobility between rows, and hence cross-pollination, and the
proportion of their flower visits as topworkers. The immediate ram-
ifications of our findings to agriculture are that pollination services
are improved by the addition of BBs to orchards pollinated by HBs.
But on a more general level, we provide strong evidence for a syn-
 G. Sapir et al. / Scientia Horticulturae 219 (2017) 107–117
ergistic effect between different bee species, mediated by resource
competition, which improves cross-pollination. This is in contra-
diction to the fear of negative interactions that might arise from
adding a competing pollinator, disrupting the pollination activity
of the main pollinator. The potential for such positive synergistic
effects should be tested in additional deciduous and subtropical
crops, and with a variety of pollinators whose presence can be
augmented, such as Osmia species (Brittain et al., 2013).
Funding
This study was supported by the Jewish National Fund (KKL) and
the Israeli Ministry of Agriculture.
References
Albrecht, M., Schmid, B., Hautier, Y., Müller, C.B., 2012. Diverse pollinator
communities enhance plant reproductive success. Proc. R. Soc. B 279,
4845–4852, http://dx.doi.org/10.1098/rspb.2012.1621.
Brittain, C., Williams, N., Kremen, C., Klein, A.M., 2013. Synergistic effects of
non-Apis bees and honey bees for pollination services. Proc. R. Soc. B 280,
20122767.
Corbet, S.A., Fussell, M., Ake, R., Frazer, A., Gunson, C., Savage, A., Smith, K., 1993.
Temperature and the pollinating activity of social bees. Ecol. Entomol. 18,
17–30.
DeGrandi-Hoffman, G., Hoopingarner, R., Baker, K.K., 1985. The influence of honey
bee ‘Sideworking’ behaviour on cross-pollination and fruit set in apple.
HortScience 20, 397–399.
Delaplane, K.S., Mayer, D.F., 2000. Crop Pollination by Bees. CABI Publication,
Wallingford, U.K.
Dennis Jr., F.G., 1979. Factors affecting yield in apple with the emphasis on
‘Delicious’. Horticult. Rev. 1, 395–422.
Dennis Jr., F.G., 2003. Flowering, pollination and fruit set and development. In:
Ferree, D.C., Wallington, I.J. (Eds.), Apples: Botany, Production and Uses. CABI
Publishing, Warrington, UK, pp. 153–166.
Dimou, M., Taraza, S., Thrasyvoulou, A., Vasilakakis, M., 2008. Effects of bumble bee
pollination on greenhouse strawberry production. J. Apicult. Res. 47, 99–101.
Drezner-Levy, T., Shafir, S., 2007. Parameters of variable reward distributions that
affect risk sensitivity of honey bees. J. Exp. Biol. 210, 269–277.
Free, J.B., Williams, I.H., 1972. The transport of pollen on the body hair of honeybees
(Apis mellifera L.) and bumblebees (Bombus spp. L.). J. Appl. Ecol. 9, 609–615.
Free, J.B., 1993. Insect Pollination of Crops, 2nd ed. Academic Press, London, UK.
Goldway, M., Sapir, G., Stern, R.A., 2007. Molecular basis and horticultural
application of the gametophytic self-incompatibility system in Rosaceae tree
fruits. In: Janick, J. (Ed.), Plant Breeding Reviews, vol. 28. John Wiley and Sons,
Inc. Publishing, USA, pp. 215–237 (chapter 7).
Goodell, K., Thomson, J.D., 1997. Comparisons of pollen removal and deposition by
honey bees and bumble bees visiting apple. Acta Hortic. 437, 103–107.
Goulson, D., 2010. Bumblebees: Behavior, Ecology and Conservation, 2nd ed.
Oxford University Press Inc., New York, NY, USA.
Greenleaf, S.S., Kremen, C., 2006. Wild bees enhance honey bees’ pollination of
hybrid sunflower. Proc. Natl. Acad. Sci. U. S. A. 103 (13), 890–913, http://dx.doi.
org/10.1073/pnas.0600929103.
Heinrich, B., 2004. Bumblebee Economics. Harvard University Press, Cambridge,
MA, USA.
Hoopingarner, R.A., Waller, G.D., 1993. Crop pollination. In: Graham, J.M. (Ed.), The
Hive and the Honeybee. Dadant and Sons, Carthage, Illinois, USA, pp.
1044–1080.
Inouye, D.W., 1978. Resource partitioning in bumblebees: experimental studies of
foraging behavior. Ecology 59, 672–678, http://dx.doi.org/10.2307/1938769.
Ish-Am, G., 2009. Directed honeybees: a preliminary investigation of an innovative
solution for passionfruit pollination in Israel. Isr. J. Plant Sci. 57, 243–251,
http://dx.doi.org/10.1560/ijps.57.3.243.
Jacquemart, A.L., Michotte-Van Der Aa, A., Raspe, Q., 2006. Compatibility and
pollinator efficiency tests on Pyrus communis L. cv. ‘Conference’. J. Hortic. Sci.
Biotechnol. 81, 827–830.
Kearns, C.A., Inouye, W.D., 1997. Pollinators, flowering plants, and conservation
biology. Bioscience 47, 297–307.
Klein, A.M., Vaissière, B.E., Cane, J.H., Steffan-Dewenter, I., Cunningham, S.A.,
Kremen, C., Tscharntke, T., 2007. Importance of pollinators in changing
landscapes for world crops. Proc. R. Soc. B 274, 303–313, http://dx.doi.org/10.
1098/rspb.2006.3721.
Klein, A.M., Brittain, C., Hendrix, S.D., Thorp, R., Williams, N., Kremen, C., 2012.
Wild pollination services to California almond rely on semi-natural habitat. J.
Appl. Ecol. 49, 723–732, http://dx.doi.org/10.1111/j.1365-2664.2012.02144.x.
Kuhn, E.D., Ambrose, T., 1982. Foraging behaviour of honey bees on ‘Golden
Delicious’ and ‘Delicious’ apple. J. Am. Soc. Hortic. Sci. 107, 391–395.
Lundberg, H., 1980. Effects of weather on foraging-flights of bumblebees
(Hymenoptera, Apidae) in a subalpine/alpine area. Holarct. Ecol. 3, 104–110.
117
Mayer, D.F., 1984. Behaviour of pollinators on Malus. In: 5th Int. Pollination Symp.,
Versailles, 1983, pp. 387–390.
Pressman, E., Shaked, R., Rosenfeld, K., Hefets, A., 1999. Study of the efficiency of
bumble bees and an electric bee in pollinated unheated greenhouse tomatoes.
J. Hortic. Sci. Biotechnol. 74, 101–104.
Roberts, R.H., 1945. Blossom structure and setting of ‘Delicious’ and other apple
varieties. Proc. Am. Soc. Hortic. Sci. 46, 87–90.
Robinson, W.S., 1979. Effect of apple cultivar on foraging behaviour and pollen
transfer by honey bees. J. Am. Soc. Hortic. Sci. 104, 596–598.
Schneider, D., Stern, R.A., Eisikowitch, D., Goldway, M., 2001. Analysis of S-allele by
PCR for determination of compatibility in the Red Delicious apple orchard. J.
Hortic. Sci. Biotechnol. 76, 596–600.
Schneider, D., Stern, R.A., Eisikowitch, D., Goldway, M., 2002. The relationship
between floral structure and honeybee pollination efficiency in Jonathan and
Topred apple cultivars. J. Hortic. Sci. Biotechnol. 77, 48–51.
Schneider, D., Stern, R.A., Goldway, M., 2005. A comparison between semi- and
fully compatible apple pollinators grown under suboptimal conditions.
HortScience 40, 1280–1282.
Schneider, D., Goldway, M., Rotman, N., Adato, I., Stern, R.A., 2009. Cross
pollination improves ‘Orri’ mandarin fruit yield. Sci. Hortic. 122, 380–384.
Stern, R.A., Eisikowitch, D., Dag, A., 2001. Sequential introduction of honeybee
colonies and doubling their density increases cross-pollination, fruit set and
yield in ‘Red Delicious’ apple. J. Hortic. Sci. Biotechnol. 76, 17–23.
Stern, R.A., Goldway, M., Zisovich, A., Shafir, S., Dag, A., 2004. Sequential
introduction of honeybee colonies increases cross-pollination, fruit-set and
yield of ‘Spadona’ pear (Pyrus communis L.). J. Hortic. Sci. Biotechnol. 79,
652–658.
Stern, R.A., Sapir, G., Shafir, S., Dag, A., Goldway, M., 2007. The appropriate
management of honeybee colonies for pollination of Rosaceae fruit trees in
warm climates. Middle Eastern and Russian. J. Plant Sci. Bitechnol. (Glob. Sci.
Books) 1, 13–19.
Stern, R.A., 2015. The photosynthesis inhibitor metamitron is highly effective
thinner for ‘Golden Delicious’ apple in a warm climate. Fruits 70 (3), 127–134.
Stout, J.C., Goulson, D., 2001. The use of conspecific and interspecific scent marks
by foraging bumblebees and honeybees. Anim. Behav. 62, 183–189, http://dx.
doi.org/10.1006/anbe.2001.1729.
Tromp, J., Wertheim, S.J., 2005. Fruit growth and development. In: Tromp, J.,
Webster, A.D., Wertheim, S.J. (Eds.), Fundamentals of Temperate Zone Tree
Fruit Production. Backhuys Publishers, Leiden, The Netherlands, pp. 240–266.
Tuell, J.K., Isaacs, R., 2010. Weather during bloom affects pollination and yield of
highbush blueberry. J. Econ. Entomol. 103, 557–562.
Van den Eijnde, J., De Ruijter, A., Van der Steen, J., 1991. Methods for rearing
Bombus terrestris continuously and the production of bumble bee colonies for
pollination purposes. Acta Hortic. 288, 154–158.
Velthuis, H.H.W., Van Doorn, A., 2006. A century of advances in bumblebee
domestication and the economic and environmental aspects of its
commercialization for pollination. Apidologie 37, 421–451.
Vicens, N., Bosch, J., 2000. Weather-dependent pollinator activity in an apple
orchard, with special reference to Osmia cornuta and Apis mellifera
(Hymenoptera: Megachilidae and Apidae). Environ. Entomol. 29, 413–420,
http://dx.doi.org/10.1603/0046-225x-29.3.413.
Waddington, K.D., 1980. Flight patterns of foraging bees relaitve to density of
artificial flowers and distribution of nectar. Oecologia 44, 199–204, http://dx.
doi.org/10.1007/bf00572680.
Westwood, N.M., 1993. Temperate Zone Pomology: Physiology and Culture, 3rd ed.
Timber Press, Portland, Oregon, USA.
Willmer, P.G., Bataw, A.A.M., Highes, J.P., 1994. The superiority of bumblebees to
honeybees as pollinators: insect visits to raspberry flowers. Ecol. Entomol. 19,
271–284.
Wolf, F., Moritz, R.F.A., 2008. Foraging distance in Bombus terrestris L.
(Hymenoptera: Apidae). Apidologie 39, 419–427.
Zhang, C., Tateishi, N., Tanabe, K., 2010. Pollen density on the stigma affects
endogenous gibberellin metabolism, seed and fruit set, and fruit quality in
Pyrus pyrifolia. J. Exp. Bot. 61, 4291–4302.
Zhang, H., Huang, J., Williams, P.H., Vaissière, B.E., Zhou, Z., Gai, Q., Dong, J., 2015.
Managed bumblebees outperform honeybees in increasing peach fruit set in
china: different limiting processes with different pollinators. PLoS One, http://
dx.doi.org/10.1371/journal.pone.0121143 (March 23, 2015).
Zisovich, A., Stern, R.A., Shafir, S., Goldway, M., 2004. Identification of seven
S-alleles from the European pear (Pyrus communis) and the determination of
compatibility among cultivars. J. Hortic. Sci. Biotechnol. 79, 101–106.
Zisovich, A., Stern, R.A., Shafir, S., Goldway, M., 2005. Fertilisation efficiency of
semi- and fully-compatible European pear (Pyrus communis L.) cultivars. J.
Hortic. Sci. Biotechnol. 80, 143–146.
Zisovich, A., Raz, A., Stern, R.A., Goldway, M., 2010. Syrian pear (Pyrus syriaca) as a
pollinator for European pear (Pyrus communis) cultivars. Sci. Hortic. 125,
256–262.
Zisovich, A., Goldway, M., Schneider, D., Steinberg, S., Stern, E., Stern, R.A., 2012.
Adding bumblebees (Bombus terrestris) to pear orchards increases seed number
per fruit, fruit set, fruit size and yield. J. Hortic. Sci. Biotechnol. 87, 353–359.