ArticlePDF Available

Reproductive environment affects learning performance in bumble bees

Authors:
Article

Reproductive environment affects learning performance in bumble bees

Abstract and Figures

Despite a presumed fitness advantage for individuals with well-developed cognitive abilities, learning performance is usually found to be highly variable within a population. Although little is currently known about the mechanisms responsible for maintaining such variation, there is correlative evidence to suggest that learning performance may be linked to reproductive physiology in the social insects. Bumble bee colonies naturally undergo an initial co-operative phase, when only the queen reproduces, and a subsequent competition phase when all colony members compete to produce male offspring. We experimentally induced these distinct phases by manipulating the presence/absence of the queen and assessed changes in sucrose responsiveness and learning performance. We found that nest-based workers upregulated their reproductive potential in queenless colonies, and correspondingly, these bees were more responsive to sucrose than their queenright counterparts, performing better in an olfactory learning task as a result. These findings suggest that differences in ovarian development are responsible for at least some of the remarkable variation in learning performance that can be observed among very closely related members of social insect colonies. Significance statement Cognitive abilities are often assumed to be inherently adaptive, so the question of why individuals vary in their learning ability has received relatively little attention. Here, we focus on reproductive status as a proximate cause of variation in learning ability in the social insects. We show that a significant proportion of the surprising variation that exists between genetically similar colony members can be explained by worker ovarian development; reproductively active workers are more sensitive to food rewards and thus learn more quickly. Learning ability may be one of a suite of correlated traits that are linked to reproductive physiology in social insects and therefore play an important role in the evolution of division of labour.
This content is subject to copyright. Terms and conditions apply.
ORIGINAL ARTICLE
Reproductive environment affects learning performance
in bumble bees
Lisa J. Evans
1,2
&Nigel E. Raine
1,3
&Ellouise Leadbeater
1
Received: 29 June 2016 /Revised: 19 August 2016 /Accepted: 25 August 2016
#Springer-Verlag Berlin Heidelberg 2016
Abstract
Despite a presumed fitness advantage for individuals with
well-developed cognitive abilities, learning performance is
usually found to be highly variable within a population.
Although little is currently known about the mechanisms re-
sponsible for maintaining such variation, there is correlative
evidence to suggest that learning performance may be linked
to reproductive physiology in the social insects. Bumble bee
colonies naturally undergo an initial co-operative phase, when
only the queen reproduces, and a subsequent competition
phase when all colony members compete to produce male
offspring. We experimentally induced these distinct phases
by manipulating the presence/absence of the queen and
assessed changes in sucrose responsiveness and learning per-
formance. We found that nest-based workers upregulated their
reproductive potential in queenless colonies, and correspond-
ingly, these bees were more responsive to sucrose than their
queenright counterparts, performing better in an olfactory
learning task as a result. These findings suggest that differ-
ences in ovarian development are responsible for at least some
of the remarkable variation in learning performance that can
be observed among very closely related members of social
insect colonies.
Significance statement
Cognitive abilities are often assumed to be inherently adap-
tive, so the question of why individuals vary in their learning
ability has received relatively little attention. Here, we focus
on reproductive status as a proximate cause of variation in
learning ability in the social insects. We show that a significant
proportion of the surprising variation that exists between ge-
netically similar colony members can be explained by worker
ovarian development; reproductively active workers are more
sensitive to food rewards and thus learn more quickly.
Learning ability may be one of a suite of correlated traits that
are linked to reproductive physiology in social insects and
therefore play an important role in the evolution of division
of labour.
Keywords Bumblebee Bombus terrestris .Foraging
behaviour .Odour learning .Reproductive environment .
Reproductive groundplan hypothesis .Sucr ose responsiveness
Introduction
Animal cognitive abilities have generally been assumed to be
associated with enhanced fitness, yet we see striking variation
in cognitive traits within species, even in societies of closely
related individuals. For example, within a single eusocial in-
sect colony there are often individuals that consistently per-
form exceedingly well in particular learning tasks while others
perform poorly (e.g. Scheiner et al. 1999,2003; Raine et al.
2006a; Raine and Chittka 2012; Perez et al. 2013; Evans and
Raine 2014a; Jandt et al. 2014). Although such variation is
Communicated by R. F. A. Moritz
Electronic supplementary material The online version of this article
(doi:10.1007/s00265-016-2209-9) contains supplementary material,
which is available to authorized users.
*Lisa J. Evans
lisa.evans@plantandfood.co.nz
1
School of Biological Sciences, Royal Holloway University of
London, Egham, Surrey TW20 0EX, UK
2
The New Zealand Institute for Plant and Food Research,
Hamilton 3240, New Zealand
3
School of Environmental Sciences, University of Guelph, N1G 2W1,
Guelph, ON, Canada
Behav Ecol Sociobiol
DOI 10.1007/s00265-016-2209-9
well-documented, both its ultimate function and proximate
causes are still poorly understood (Jeanson and
Weid en l ler 2014). Here, we focus on a hypothesised prox-
imate cause of variation in learning ability in the social insects
using the bumble bee (Bombus terrestris) as our experimental
model organism. We investigate whether individual learning
abilities could be one of a suite of correlated traits that are
closely linked to reproductive physiology.
Like many pollinators, foraging bumble bees are faced with
a complex environment in which floral rewards can change
quickly over short time periods (Raine et al. 2006b). The
ability to learn about floral characteristics that predict high-
quality food rewards should enhance foraging efficiency, and
correspondingly, learning performance should correlate with
foraging success (Raine and Chittka 2008). Nonetheless, in-
dividual foragers vary markedly in their performance in
sucrose-rewarded learning tasks (Raine et al. 2006a;Muller
and Chittka 2012; Raine and Chittka 2012; Evans and Raine
2014a; Smith and Raine 2014). Studies from other social in-
sects provide reason to hypothesise that this variability could
derive from variation in reproductive physiology among
nestmates, because reproductive physiology correlates with a
suite of behavioural traits in social bees.
Perhaps of greatest interest here is a large body of evidence
linking both learning and reproductive status to pollen/nectar
foraging preferences within the worker caste of the bumble
bees close relative, the honey bee Apis mellifera. For exam-
ple, high ovariole numbers in young honey bee workers are
associated with a subsequent foraging preference for pollen
(Amdam et al. 2006;Nelsonetal.2007; Page and Amdam
2007; Linksvayer et al. 2009;Wangetal.2010). Likewise,
high vitellogenin (a yolk-precursor protein) and juvenile hor-
mone (JH) titres in workers are associated with pollen-
hoarding (Amdam et al. 2004; Schultz et al. 2004), and knock-
down of the gene responsible for vitellogenin production re-
sults in a subsequent preference for nectar collection (Nelson
et al. 2007). We hypothesise that variation in worker repro-
ductive physiology might have impacts upon learning because
honey bee preferences for pollen or nectar collection are also
associated with their responsiveness to sucrose (Pankiw and
Page 1999; Scheiner et al. 1999;Scheiner2004; Latshaw and
Smith 2005; Drezner-Levy et al. 2009) and correspondingly
with their performance in sucrose-rewarded learning tasks
(Scheiner et al. 2001a,2001b). Nectar-specialist honey bee
workers have lower sucrose responsiveness (SR) and conse-
quently collect more concentrated nectar (because they are
less likely to detect or respond to more dilute concentrations).
Although reproductive traits are linked to nectar/pollen
preferences, and nectar/pollen preferences are in turn linked
to sucrose-rewarded learning, the evidence for a link between
reproductive traits and sucrose-rewarded learning remains
correlative and indirect. In this study, we exploit a natural
feature of the bumble bee life cycle to manipulate worker
reproductive development and directly assay the conse-
quences for sucrose responsiveness and odour learning perfor-
mance. Towards the end of the annual bumble bee colony
cycle, previously co-operative workers enter an all-out com-
petition to produce male offspring (Van Honk and Hogeweg
1981; Bourke and Ratnieks 2001), which continues during
colony decline and eventual death.In this study, we artificially
induced thisreproductivedevelopmentby splitting colonies to
create matched queenright and queenless groups. We verified
that our manipulation led to reproductive activation in
queenless workers and assayed differences between the two
sub-colony types in terms of sucrose responsiveness and
learning performance using the proboscis extension reflex
(PER) conditioning paradigm.
Methods
Colony setup
Five young B. terrestris colonies, containing an average of 22
workers, were obtained from Biobest (Westerlo, Belgium). On
arrival, each colony was transferred into a colony box
(L × W × H = 280 × 155 × 110 mm). Colony boxes were
connected, via rubber tubing, to individual foraging arenas
(235 × 140 × 110 mm) containing a dish of defrosted, honey
bee-collected pollen (Koppert UK Ltd; ca. 3 g replaced every
second day) and ad libitum 40 % (v/v) sucrose solution in a
feeder. All workers that emerged prior to splitting colonies
(described below) were individually marked with numbered
tags (Opalith tags; Christian Graze KG, Germany) on the day
of emergence, so that their subsequent age when tested was
known. Colonies were split when they contained 75100
workers and were still in their Bpre-competition phase^, when
all egg laying is exclusive to the queen. Each colony was split
to create one queenright (Q+) group and one queenless (Q)
group: both the bees and brood were divided evenly, and half
of them were transferred to a new colony box. In small groups
of queenless workers ovary development takes just 5 days
(R seler 1974), after which time workers of seven (or more)
days of age can be ready to lay eggs (Amsalem et al. 2014).
Seven to 10 days after splitting the colonies, we assessed the
sucrose responsiveness and olfactory learning performance of
workers from both queenless and queenright sides.
Behavioural observations of workers
We expected that any effect of the queenless state might vary
with worker task specialization (nest workers have more de-
veloped ovaries in some bee species: West-Eberhard 1996)
and/or worker dominance (dominant workers may have more
developed ovaries: Amsalem and Hefetz 2010; Amsalem et al.
2013). To control for these variables in our statistical analysis
Behav Ecol Sociobiol
we monitored individual worker behaviour both before and
after colony splits. Activity in the foraging box was observed
for 11.5hperday(23 × 30-min intervals spread over the
course of the day), 5 days per week. Foragers were defined as
individuals that were observed foraging at one of the feeders
on at least five occasions prior to splitting (as colonies were
initially small they needed less pollen and sucrose solution) or
eight times (range = 830) after splitting. Nest workers were
defined as individuals observed on feeders on no more than
four (range = 04) occasions. These simple definitions of
Bnest worker^and Bforager^were used to increase the power
of our analysis by accounting for task specialization (role)a
known correlate of variation in reproductive development
among workers.
To identify individuals displaying dominant behaviour, we
performed focal observations of each queenless nest (no dom-
inant behaviours were observed in queenright colonies) under
red light twice daily for 15 min. Workers were considered to
be dominant if they displayed any of the following behav-
iours: overt aggression, threatening behaviour, oviposition
and/or egg eating (Van Doorn and Heringa 1986; Duchateau
and Velthuis 1988; Bloch et al. 2000;Gevaetal.2005). When
possible, the behaviours of aggressed individuals (workers
receiving the aggressive behaviour) were also recorded, be-
cause competing workers are reported to focus their aggres-
sion on those workers signalling fertility (Van der Blom and
Verkade 1991; Visscher and Dukas 1995; Amsalem et al.
2009). When selecting foragers and nest workers from the
queenless treatment groups for learning and sucrose respon-
siveness assays, we selected bees that performed or received
aggressive behaviours. In other words, we chose to test those
individuals most likely to have responded to the queenless
state.
Sucrose responsiveness
We assessed the sucrose responsiveness (SR) of workers 7
10 days after the colonies were split using a PER assay.
Between two and three foragers and two to three nest bees
were caught from each treatment group per day and chilled
on ice until they became quiescent (ca.1215 min). They
were then mounted in modified plastic syringe tubes (2 ml
volume, 8.2 mm diameter, cut down to 3 cm long), with a
notch cut out of the top of the tube to allow the bees to freely
extend their proboscis. Subjectsheads were held in place
between two entomological pins (size 0) that were glued at
one end and subsequently fastened at the other end. This pin
yoke was secured with tape. A small ball of moistened cotton
wool was inserted into the mounting tube behind the bee to
prevent desiccation (Riveros and Gronenberg 2009; Smith
and Raine 2014). The harnesses were individually numbered
to ensure that testing was conducted blind with respect to
treatment group.
Each subjects sucrose responsiveness was assessed 23h
after being harnessed, using an ascending concentration series
of 0 (water; negative control), 1, 3, 5, 8, 10, 15, 25, 35 and
50 % (v/v) sucrose solution. Stimulus presentation involved
touching the subjects antennae with one of the sucrose solu-
tions, presented on a toothpick. Stimulation with detectable
levels of sucrose evokes proboscis extension (Bitterman
et al. 1983), with water included as a negative control (no
subjects responded to the water solution). The time between
stimulus presentations was 45 min, depending on the number
of bees (n=2
130) being tested. An individualsSRscore
was the number of concentrations that evoked a full extension
of the proboscis (010), with a high SR score indicating great-
er responsiveness to sucrose.
Olfactory learning performance
After assaying sucrose responsiveness, bees were left over-
night to ensure that they were sufficiently motivated for as-
sessment of olfactory learning performance by PER the fol-
lowing morning. This period (~18 h) is necessary to prevent
bumble bees from ceasing to respond during PER testing. We
used an absolute PER conditioning paradigm (Riveros and
Gronenberg 2009; Smith and Raine 2014). Prior to testing,
each bees antennae were stimulated with a 50 % (v/v) sucrose
solution, and if the bee extended its proboscis it was given a
small droplet of sucrose. If a bee failed to respond after four
attempts it was excluded from the test (n= 18 of 172 bees,
including 10 bees that died overnight).
Bees were assessed on their ability to learn to associate a
floral odour (lavender essential oil, Calmer Solutions) with a
50 % sucrose solution reward. Each harnessed bee was placed
individually into an odour extraction hood during a training
event. The bee was positioned in a marked location, 3 cm
away from an odour tube that delivered an air stream (contain-
ing the odour stimulus at the appropriate times). The odour
tube contained a strip of filter paper soaked with 2 μlofthe
essential oil. This concentration is similar to those used in
other studies of olfactory learning in bumble bees (e.g.
Riveros and Gronenberg 2009; Smith and Raine 2014)and
is most likely high enough to overwhelm any individual dif-
ferences in odour sensitivity between bees, which could oth-
erwise affect learning (Spaethe et al. 2007). Filter paper strips
were replaced after 2030 uses for consistency of odour
strength. A programmable logic controller computer con-
trolled the volume of the air, flow rate and duration of the
odour presentation. Bees were exposed to 5 s of clean air,
followed by 10 s of odour-containing air for each trial.
Using a Gilmont syringe, each bee was provided with 0.8 μl
of sucrose solution after approximately 6 s of exposure to the
odour-containing air.
A positive response was recorded in trials when a bee ex-
tended its proboscis during exposure to the odour stimulus
Behav Ecol Sociobiol
prior to antennal stimulation with sucrose solution. On such
occasions, the sucrose reward was provided immediately.
Each bee was given 15 trials, with a 12-min inter-trial interval.
After each trial, we recorded whether the bee responded be-
fore antennal sucrose stimulation (demonstrating an associa-
tion between odour and reward), after stimulation (demon-
strating a response to antennal stimulation with sucrose) or
no response. Learning performance was quantified as the
sum of trials in which an individual bee extended its proboscis
prior to antennal stimulation. Bees that did not respond to the
odour, or the subsequent antennal sucrose stimulation, for
three consecutive trials were excluded from the learning ex-
periment (n= 9 bees), as were unresponsive bees during the
SR test (n=10bees).
Bees were immediately frozen at 20 °C after testing.
Thorax width measurements were taken for each tested individ-
ual and recorded as a measure of worker body size, which has
the potential to affect learning performance (Worden et al.
2005; Sommerlandt et al. 2014; but see Raine et al. 2006a;
Raine and Chittka 2008; Evans and Raine 2014b; Smith and
Raine 2014). Workers were then dissected to assess the pres-
ence of measurable oocytes in each of the eight (four per ovary)
ovarioles (egg-laying filaments).Whennoclearlydefinedoo-
cytes were present in the ovarioles, it was recorded that the
ovarioles were present but undeveloped. Note that Bovarian
development^was measured as a general indicator of reproduc-
tive activation, but it is just one component of the large network
of physiological and endocrine reproductive traits that are likely
to change following queen removal. Our analyses focus upon
the effects of the queenless/queenright state, rather than specif-
ically on the effects of ovarian development.
Statistical analyses
All analyses were conducted in R v 3.0.2 (R Core
Development Team 2014). We used a series of generalised
linear mixed models, GLMMs (using the glmer function in
package lme4: Bates et al. 2014, assuming a binomial error
distribution) to examine whether either SR (Table 1(a)) and/or
olfactory learning performance (Table 1(b)) were predicted by
treatment (queenright/queenless). In both cases, we used an
information theory approach to compare a series of candidate
models containing all combinations of treatment (queenright/
queenless), worker size, worker age and role (task specializa-
tion: nest worker/forager) as predictors (R package MuMIn:
Bartoń2016). We also included an interaction term
(role:treatment) because initial data exploration indicated that
the effects of treatment varied between the forager and nest
worker groups. Each candidate model included Bcolony^as a
random factor, including our null model (Bbasic model^),
which otherwise contained only the intercept as a predictor.
Parameter estimates are based on model averaging of the 95 %
confidence set.
Results
Overall, we found that both SR and olfactory learning perfor-
mance varied significantly between the queenright and
queenless groups, and that this effect was limited to the nest
worker bees (Figs. 1and S1). For SR, model-averaged param-
eter estimates based on the 99 % confidence set of models
(Table 1(a)) indicated that treatment (model-averaged estimate
1.46, 95 % C.I. = 0.69 to 2.23), role (0.69; 0.34 to 1.04) and an
interaction between treatment and role (1.47; 1.99 to 0.95)
significantly predicted SR. Worker age and worker size did
not receive strong support (estimates and 95 % C.I. = 0.02;
0.004to0.05and0.05; 0.42 to 0.31 respectively).
Likewise, treatment (estimate 2.38; 1.66 to 3.10), role (esti-
mate 0.51; 0.22 to 0.80) and their interaction (estimate 1.93;
2.44 to 1.42) emerged as significant predictors of olfactory
learning performance, and in this case both worker size (esti-
mate 1.55; 1.18 to 1.92) and worker age (estimate 0.03;
0.06 to 0.01) also received support (Table 1(b)). In summa-
ry, nest workers in queenless colonies were more responsive
to sucrose and thus performed better in the olfactory learning
task than their counterparts in the queenright colonies, but no
difference was found among foragers.
Post-experimental dissections revealed that this marked
difference between young nest workers and older foragers
most likely reflects the differential effects of queen removal
on the two worker types. In the presence of the queen, nest
workers typically had undeveloped ovaries, but queen remov-
al led to ovarian development. In contrast, forager ovaries
were well developed in both queenless and queenright colo-
nies (Fig. 2). However, replacing Btreatment^with Bovarian
development^in our full models did not provide a better fit to
the SR data (AICc [treatment model] = 1059.3, AICc [ovarian
development model] = 1111.3), nor to our learning perfor-
mance data (AICc [treatment model] = 948.0, AICc [ovarian
development model] = 961.2). This most likely reflects the
fact that variation in ovarian development, at least when
categorised using our methods, is unlikely to capture all the
reproductive changes that occur in workers when the queen is
removed from a colony.
Discussion
In this study, we found that queenless bumble bee workers
were more responsive to sucrose than queenright (control)
bees. Correspondingly, they performed better in an olfactory
learning task. Interestingly, this effect is limited to those bees
that specialised upon nest-based tasks, rather than those bees
leaving the colony to forage. Our ovarian dissections suggest
an explanation for this: in a queenright colony, nest workers
show less ovarian development than foragers, but this in-
creases to forager-like levels following removal of the queen.
Behav Ecol Sociobiol
This indicates that effects of queen removal may be greater for
nest-based workers on the network of endocrine and physio-
logical traits that underlie reproduction. In other words, for
those individuals that switched to a reproductive phenotype
upon removal of the queen, we saw a corresponding change in
SR and learning ability.
Ultimately, why should workers benefit from increased re-
sponsiveness to sucrose in the absence of a queen?
Reproductive workers could be under selection to participate
in foraging, to provide for their brood, but there is no obvious
direct link between increased responsiveness to sucrose and
foraging performance. Greater responsiveness is not likely to
lead to collection of higher-quality nectar; in contrast, bees
that are very responsive to sucrose should detect and accept
less concentrated (and thus poorer quality) floral nectar re-
wards (Pankiw et al. 2001). Alternatively, it may be that the
increase in sucrose responsiveness that we observed is a by-
product of selection for responsiveness to other stimuli in
reproductive workers, because there is evidence to suggest
that heightened sucrose responsiveness reflects a general in-
crease in responsiveness to many stimuli (e.g. light; Erber
et al. 2006; Tsuruda and Page 2009). For example, honey
Tabl e 1 Candidate models to predict the (a) sucrose responsiveness (n= 157 bees) and (b) olfactory learning performance (n= 144 bees) for
B. terrestris workers
Age Role Treatment Size Treatment:role AICc ΔAICc Akaike weight Cumulative weight
(a)
Yes Yes Yes Yes 929.00 0.00 0.41 0.41
Yes Ye s Ye s 9 2 9 .4 0 0.4 0 0 . 33 0.7 4
Yes Yes Yes Yes Yes 930.90 1.95 0.15 0.89
Yes Yes Yes Yes 931.50 2.57 0.11 1.00
Yes Yes 955.00 26.00 0.00 1.00
Yes Yes Yes 956.90 27.96 0.00 1.00
Yes Yes Yes 957.10 28.09 0.00 1.00
Yes Yes Yes Yes 959.10 30.09 0.00 1.00
Yes 968.00 39.05 0.00 1.00
Yes Yes 969.10 40.12 0.00 1.00
Yes Yes 969.20 40.22 0.00 1.00
Yes Yes Yes 970.70 41.71 0.00 1.00
Yes 982.90 53.90 0.00 1.00
Yes Yes 984.80 55.82 0.00 1.00
Yes Yes 985.00 55.98 0.00 1.00
Yes Yes Yes 986.90 57.92 0.00 1.00
Yes 996.60 67.61 0.00 1.00
998.20 69.24 0.00 1.00
Yes Yes 998.40 69.42 0.00 1.00
Yes 999.30 70.36 0.00 1.00
(b)
Yes Yes Yes Yes Yes 1017.0 0.00 0.93 0.93
Yes Yes Yes Yes 1022.2 5.22 0.07 1.00
Yes Yes 1076.4 59.48 0.00 1.00
Yes Yes Yes 1076.5 59.57 0.00 1.00
Yes Yes Yes 1078.6 61.62 0.00 1.00
Yes Yes Yes Yes 1078.6 61.65 0.00 1.00
Yes 1079.2 62.24 0.00 1.00
Yes Yes 1080.2 63.22 0.00 1.00
Yes Yes 1081.3 64.36 0.00 1.00
Yes Yes Yes 1082.2 65.28 0.00 1.00
Yes Yes Yes 1093.2 76.21 0.00 1.00
Yes Yes Yes Yes 1095.0 78.06 0.00 1.00
Yes Yes Yes 1154.4 137.40 0.00 1.00
Yes Yes 1156.5 139.58 0.00 1.00
Yes Yes 1161.7 144.78 0.00 1.00
Yes Yes 1168.8 151.83 0.00 1.00
Yes 1169.3 152.39 0.00 1.00
Yes 1171.7 154.76 0.00 1.00
Yes 1173.6 156.63 0.00 1.00
1182.0 165.09 0.00 1.00
Each model contained the intercept and Bcolony^as a random factor, together with fixed factors indicated by the column headings. The 95 % confidence
set, shown in bold, is the set of models in which we can at least be 95 % sure that the best model lies
Behav Ecol Sociobiol
bee larvae produce a Bbrood pheromone^to stimulate pollen
foraging (Pankiw et al. 1998), and it may be that reproductive
bumble bee workers respond to a similar substance.
A third alternative is that increased sucrose responsiveness
in reproductive workers has no adaptive function. The repro-
ductive ground plan hypothesis postulates that reproductive
behaviour, foraging preferences and neurosensory responses
are linked by a pleiotropic gene network that simply occurred
at the same life-history stage in a solitary ancestor, and have
been co-selected to produce division of foraging labour in
extant social insects (Amdam et al. 2004,2006; Page and
Amdam 2007). Under this framework, the increase in sucrose
responsiveness and corresponding improvement in learning
ability that we observed could be simply a by-product of re-
productive upregulation.
Our findings reveal novel parallels with other
Hymenoptera in terms of social biology. We found that only
nest-based workers responded to the queenless state with re-
productive activation. In honey bees (A. mellifera), it is the
younger, nest-based workers that are also more likely (than
older foragers) to undergo oogenesis in the absence of the
queen (Lin et al. 1999), just as we observed for bumble bees.
Likewise, in the tropical wasps Metapolybia aztecoides and
Synoeca surinama (both of which exhibit cyclical oligyny) a
dead or missing queen is typically replaced by recently
emerged females that were already present in the nest (i.e. they
have not yet begun to forage). In the absence of a queen, these
nest-dwelling individuals will mate and develop mature ova-
ries (West-Eberhard 1996). However, there are also clear dif-
ferences between the reproductive systems of these species
and B. terrestris. In queenright M. aztecoides,S. surinama
and A. mellifera, foraging individuals have regressed ovaries
(West-Eberhard 1996; Pinto et al. 2000), while we found that
queenright B. terrestris foragers exhibited high levels of ovar-
ian development. This may reflect the fact that ovarian devel-
opment in bumble bee colonies is actively suppressed through
the presence of the queen, through queen mandibular phero-
mone (QMP) and brood pheromone (Trynor et al. 2014).
Since foragers spend much less of their time in the nest, they
are likely to receive lower exposure to both pheromones,
allowing for initiation of JH secretion by the corpora allata
(R seler 1977). However, it is not clear why the same effect is
not observed in other eusocial insects, since the mechanisms
by which worker reproduction is suppressed are similar. One
possible explanation stems from the fact that worker bumble
bees are almost unique among social insects for their lack of
age polyethism (Cameron 1989; Cameron and Robinson
1990). Throughout the social insects, it is typically the
youngest workers that develop ovaries in the absence of
the queen, but only in bumble bees is it likely that some
of these young workers will regularly leave the nest to
forage. Alternatively, it is possible that in our experimen-
tal setup (foragers collecting food in a flight arena), for-
agers still had sufficient energy for ovarian development,
which would perhaps not occur under natural conditions
when foraging from flowers in the field (Foster et al.
2004; Williams et al. 2008; Jandt and Dornhaus 2011).
In summary, our results provide evidence that at least some
of the appreciable intracolony variation seen in learning per-
formance among bumble bees (Raine et al. 2006a;Mullerand
528
0
1
2
3
4
5
6
Mean number of learnt responses
Treatment
b
FQ+ FQ- NQ+ NQ-
0
1
2
3
4
5
6
7
8
9
Mean number of responses in SR
assesment
Treatment
a
FQ+ FQ- NQ+ NQ-
a
a
a
b
a
a
a
b
Fig. 1 a Sucrose responsiveness (number of sucrose concentrations that
evoked full proboscis extension; a high SR score indicates high
responsiveness to sucrose) in queenright (Q+: darker-coloured bars)
compared to queenless treatment groups (Q:lighter-coloured bars)
andinforagers(F:red bars) compared to nest workers (N: purple
bars). bLearning performance in queenright (darker-coloured bars)
compared to queenless treatment groups (lighter-coloured bars)andin
foragers (red bars) compared to nest workers (purple bars). Column
heights indicate the mean (± SE) values for SR (a) and learning
performance (b) of bees from each group. Column heights indicate the
mean (± SE) sucrose responsiveness (a) and learning performance (b)of
bees from each group. Different letters indicate significant differences
among groups (TukeysHSD,p0.05)
0
10
20
30
40
50
60
70
80
90
% of bees showing some ovarian
development
FQ+ FQ- NQ+ NQ-
Fig. 2 Ovarian development in queenright (Q+: darker-coloured bars)
compared to queenless treatment groups (Q:lighter-coloured bars)and
in foragers (F: red bars) compared to nest workers (N: purple bars)
Behav Ecol Sociobiol
Chittka 2012; Evans and Raine 2014a;SmithandRaine2014;
Sommerlandt et al. 2014) is likely to be attributable to indi-
vidual variation in reproductive development. In queenright
colonies, we see variation in ovarian development between
nest workers and foragers, and correlated variation in sucrose
responsiveness. Our results also lead to the testable prediction
that in bumble bee colonies that have reached the
Bcompetition phase^, whereby the queen loses reproductive
dominance, we may see similarly high levels of reproductive
development between nest workers and foragers, and thus less
variation in responsiveness to sucrose and learning behaviour
than in early-stage colonies. On a proximate level, our study
provides strong evidence for a link between reproductive
physiology and cognitive traits in bees, inviting further inves-
tigation into the potential fitness benefits of this relationship.
Acknowledgments
We thank Gemma Baron, Mark Brown, Karen Smith and Dara
Stanley for useful discussions and David Pattemore for his
comments on an earlier version of the manuscript. LE was
supported by Plant and Food Research New Zealand, a C.
Alma Baker PhD Scholarship, an A.G.K. Overseas Research
Scholarship and a Queen Elizabeths Technicians Award dur-
ing this work. EL is supported by a Leverhulme Trust Early
Career Fellowship. NR is supported as the Rebanks Family
Chair in Pollinator Conservation by The W. Garfield Weston
Foundation.
Compliance with ethical standards
Author contributions EL and LJE conceived the project. LJE, EL and
NER designed the research experiments. LJE carried out the experiments.
LJE and EL performed the statistical analyses. LJE, EL and NER wrote
the manuscript.
References
Amdam GV, Norberg K, Fondrk MK, Page RE (2004) Reproductive
ground plan may mediate colony-level selection effects on individ-
ual foraging behavior in honey bees. Proc Natl Acad Sci U S A 101:
1135011355. doi:10.1073/pnas.0403073101
Amdam GV, Csondes A, Fondrk MK, Page RE (2006) Complex social
behaviour derived from maternal reproductive traits. Nature 439:76
78. doi:10.1038/nature04340
Amsalem E, Hefetz A (2010) The appeasement effect of sterility signal-
ing in dominance contests among Bombus terrestris workers. Behav
Ecol Sociobol 64:16851694. doi:10.1007/s00265-010-0982-4
Amsalem E, Twele R, Francke W, Hefetz A (2009) Reproductive com-
petition in the bumble bee Bombus terrestris: do workers advertise
sterility? Proc R Soc Lond B 276:12951304. doi:10.1098
/rspb.2008.1688
Amsalem E, Shamia D, Hefetz A (2013) Aggression or ovarian develop-
ment as determinants of reproductive dominance in Bombus
terrestris: interpretation using a simulation model. Insect Soc 60:
213222. doi:10.1007/s00040-013-0285-7
Amsalem E, Malka O, Grozinger C, Hefetz A (2014) Exploring the role
of juvenile hormone and vitellogenin in reproduction and social
behavior in bumble bees. BMC Evol Biol 14:45. doi:10.1186
/1471-2148-14-45
BartońK (2016) MuMIn: Multi-Model Inference. R package version
1.15.6. Retrieved from http://CRAN.R-project.org/package=
MuMIn
Bates D, Maechler M, Bolker B, Walker S (2014) lme4: Linear mixed-
effects models using Eigen and s4. R package version 1.16:
http://CRAN.R-project.org/package=lme4.
Bitterman ME, Menzel R, Fietz A, Schafer S (1983) Classical condition-
ing of proboscis extension in honeybees (Apis mellifera. J Comp 97:
107119
Bloch G, Hefetz A, Hartfelder K (2000) Ecdysteroid titer, ovary status,
and dominance in adult worker and queen bumble bees (Bombus
terrestris. J Insect Physiol 46:10331040. doi:10.1016/S0022-1910
(99)00214-0
Bourke AFG, Ratnieks FLW (2001) Kin-selected conflict in the bumble-
bee Bombus terrestris (Hymenoptera: Apidae). Proc R Soc Lond B
268:347355. doi: 10.1098/rspb.2000.1381
Cameron SA (1989) Temporal patterns of division of labor among
workers inthe primitively eusocial bumble bee, Bombus griseocollis
(Hymenoptera: Apidae). Ethology 80:137151. doi:10.1111/j.1439-
0310.1989.tb00735.x
Cameron SA, Robinson GE (1990) Juvenile hormone does not affect
division of labor in bumble bee colonies (Hymenoptera: Apidae).
Ann Entomol Soc Am 83:626631. doi:10.1093/aesa/83.3.626
Drezner-Levy T, Smith BH, Shafir S (2009) The effect of foraging spe-
cialization on various learning tasks in the honey bee (Apis
mellifera). Behav Ecol Sociobol 64:135148. doi:10.1007/s00265-
009-0829-z
Duchateau MJ, Velthuis HHW (1988) Development and reproductive
strategies in Bombus colonies. Behaviour 107:186207
Erber J, Hoormann J, Scheiner R (2006) Phototactic behaviour correlates
with gustatory responsiveness in honey bees (Apis mellifera L.).
Behav Brain Res 174:174180. doi:10.1016/j.bbr.2006.07.023
Evans LJ, Raine NE (2014a) Changes in learning and foraging behaviour
within developing bumble bee (Bombus terrestris) colonies. PLoS
One 9:e90556. doi:10.1371/journal.pone.0090556
Evans LJ, Raine NE (2014b) Foraging errors play a role in resource
exploration by bumble bees (Bombus terrestris). J Comp Physiol
A 200:475484. doi: 10.1007/s00359-014-0905-3
Foster RL, Brunskill A, Verdirame D, O'Donnell S (2004) Reproductive
physiology, dominance interactions, and division of labour among
bumble bee workers. Physiol Ento 29:327334. doi:10.1111/j.0307-
6962.2004.00388.x
Geva S, Hartfelder K, Bloch G (2005) Reproductive division of labor,
dominance, and ecdysteroid levels in hemolymph and ovary of the
bumble bee Bombus terrestris. J Insect Physiol 51:811823.
doi:10.1016/j.jinsphys.2005.03.009
Jandt JM, Dornhaus A (2011) Competition and cooperation: bumblebee
spatial organization and division of labor may affect worker repro-
duction late in life. Behav Ecol Sociobiol 65:23412349.
doi:10.1007/s00265-011-1244-9
Jandt JM, Bengston S, Pinter-Wollman N, Pruitt J, Raine NE, Dornhaus
A, Sih A (2014) Behavioral syndromes and social insects: person-
ality at multiple levels. Biol Rev 89:4867. doi:10.1111/brv.12042
Jeanson R, Weidenmüller A (2014) Interindividual variability in social
insectsproximate causes and ultimate consequences. Biol Rev 89:
671687. doi:10.1111/brv.12074
Latshaw JS, Smith BH (2005) Heritable variation in learning performance
affects foraging preferences in the honey bee (Apis mellifera.Behav
Ecol Sociobiol 58:200207. doi:10.1007/s00265-004-0904-4
Behav Ecol Sociobiol
Lin HR, Winston ML, Haunerland NH, Slessor KN (1999) Influence of
age and population size on ovarian development, and of trophallaxis
on ovarian development and vitellogenin titres of queenless worker
honey bee (Hymenoptera: Apidae). Can Entomol 13:695706.
doi:10.4039/Ent131695-5
Linksvayer TA, Rueppell O, Siegel A, Kaftanoglu O, Page RE Jr,
Amdam GV (2009) The genetic basis of transgressive ovary size
in honeybee workers. Genetics 183:693707. doi:10.1534
/genetics.109.105452
Muller H, Chittka L (2012) Consistent interindividual differences in dis-
crimination performance by bumblebees in colour, shape and odour
learning tasks (Hymenoptera: Apidae: Bombus terrestris). Entomol
Gen 34:18. doi:10.1127/entom.gen/34/2012/1
Nelson CM, Ihle KE, Fondrk MK, Page RE Jr, Amdam GV (2007) The
gene vitellogenin has multiple coordinating effects on social organi-
zation. PLoS Biol 5:e62. doi:10.1371/journal.pbio.0050062
Page RE Jr, Amdam GV (2007) The making of a social insect: develop-
mental architectures of social design. Biossays 29:334343.
doi:10.1002/bies.20549
Pankiw T, Page RE Jr (1999) The effect of genotype, age, sex, and caste
on response thresholds to sucrose and foraging behavior of honey
bees (Apis mellifera L.). J Comp Physiol A 185:207213.
doi:10.1007/s003590050379
Pankiw T, Page RE, Fondrk MK (1998) Brood pheromone stimulates
pollen foraging in honey bees (Apis mellifera). Behav Ecol
Sociobiol 44:193198. doi:10.1007/s002650050531
Pankiw T, Waddington KD, Page RE (2001) Modulation of su-
crose response thresholds in honey bees (Apis mellifera L.):
influence of genotype, feeding, and foraging experience. J
Comp Physiol A 187:293301. doi:10.1007/s003590100201
Perez M, Rolland U, Giurfa M, d'Ettorre P (2013) Sucrose responsive-
ness, learning success, and task specialization in ants. Learn
Memory 20:417420. doi: 10.1101/lm.031427.113
Pinto LZ, Bitondi MMG, Simoes ZLP (2000) Inhibition of vitellogenin
synthesis in Apis mellifera workers by a juvenile hormone analogue,
pyriproxyfen. J Insect Physiol 46:153160. doi:10.1016/S0022-
1910(99)00111-0
R Core Development Team (2014) R: a language and environment for
statistical computing. R Foundation for Statistical Computing,
Vienna URL http://www.R-project.org
Raine NE, Chittka L (2008) The correlation of learning speed and natural
foraging success in bumble-bees. Proc R Soc Lond B 275:803808.
doi:10.1098/rspb.2007.1652
Raine NE, Chittka L (2012) No trade-off between learning speed and
associative flexibility in bumblebees: a reversal learning test with
multiple colonies. PLoS One 7:e45096. doi:10.1371/journal.
pone.0045096
Raine NE, Ings TC, Ramos-Rodriguez O, Chittka L (2006a)
Intercolony variation in learning performance of a wild
British bumblebee population (Hymenoptera: Apidae:
Bombus terrestris audax). Entomol Gen 28:241256.
doi:10.1127/entom.gen/28/2006/241
Raine NE, Ings TC, Dornhaus A, Saleh N, Chittka L (2006b) Adaptation,
genetic drift, pleiotropy, and history in the evolution of bee foraging
behavior. Adv Stud Behav 36:305354. doi:10.1016/S0065-3454
(06)36007-X
Riveros AJ, Gronenberg W (2009) Olfactory learning and memory in the
bumblebee Bombus occidentalis. Naturwissenschaften 96:851856.
doi:10.1007/s00114-009-0532-y
R seler PF (1974) Comparative studies of oogenesis in bumblebee
workers (B. terrestris L) in queenright and queenless colonies.
Insect Soc 21:249274
Röseler P (1977) Juvenile hormone control of oögenesis in bumblebee
workers, Bombus terrestris. J Insect Physiol 23:985992.
doi:10.1016/0022-1910(77)90126-3
Scheiner R (2004) Responsiveness to sucrose and habituation of the pro-
boscis extension response in honey bees. J Comp Physiol A 190:
727733. doi:10.1007/s00359-004-0531-6
Scheiner R, Erber J, Page RE Jr (1999) Tactile learning and the individual
evaluation of the reward in honey bees (Apis mellifera L.). J Comp
Physiol A 185:110. doi:10.1007/s003590050360
Scheiner R, Page RE, Erber J (2001a) The effects of genotype, foraging
role, and sucrose responsiveness on the tactile learning performance
of honey bees (Apis mellifera L.). Neurobiol Learn Mem 76:138
150. doi:10.1006/nlme.2000.3996
Scheiner R, Page RE Jr, Erber J (2001b) Responsiveness to sucrose af-
fects tactile and olfactory learning in preforaging honey bees of two
genetic strains. Behav Brain Res 120:6773. doi:10.1016/S0166-
4328(00)00359-4
Scheiner R, Barnert M, Erber J (2003) Variation in water and sucrose
responsiveness during the foraging season affects proboscis exten-
sion learning in honey bees. Apidologie 34:6772. doi:10.1051
/apido:2002050
Schultz DJ, Pankiw T, Fondrk MK, Robinson GE, Page RE Jr (2004)
Comparisons of juvenile hormone hemolymph and octopamine
brain titers in honey bees (Hymenoptera: Apidae) selected for high
and low pollen hoarding. Ann Entomol Soc Am 97:13131319
Smith KE, Raine NE (2014) A comparison of visual and olfactory learn-
ing performance in the bumblebee Bombus terrestris. Behav Ecol
Sociobiol 68:15491559. doi:10.1007/s00265-014-1765-0
Sommerlandt FMJ, Roessler W, Spaethe J (2014) Elemental and non-
elemental olfactory learning using PER conditioning in the bumble-
bee, Bombus terrestris. Apidologie 45:106115. doi:10.1007
/s13592-013-0227-4
Spaethe J, Brockmann A, Halbig C, Tautz J (2007) Size determines an-
tennal sensitivity and behavioral threshold to odors in bumblebee
workers. Naturwissenschaften 94:733739. doi:10.1007/s00114-
007-0251-1
Trynor KS, Conte YL, Page RE (2014) Queen and young larvae phero-
mones impact nursing and reproductive physiology of honey bee
(Apis mellifera) workers. Behav Ecol Sociobiol 68:20592073.
doi:10.1007/s00265-014-1811-y
Tsuruda JM, Page RE Jr (2009) The effects of foraging role and genotype
on light and sucrose responsiveness in honey bees (Apis mellifera
L.). Behav Brain Res 205:132137. doi:10.1016/j.bbr.2009.07.022
Van der Blom J, Verkade AJ (1991) Does kin recognition in honey bees
Apis mellifera L. influence the workers' chances of becoming egg
layers? Anim Behav 42:867870. doi:10.1016/S0003-3472(05
)80136-8
Van Doorn A, Heringa J (1986) The ontogeny of a dominance hierarchy
in colonies of the bumblebee Bombus terrestris Hymenoptera
Apidae. Insect Soc 33:325. doi:10.1007/BF02224031
Van Honk C, Hogeweg P (1981) The ontogeny of the social structure in a
captive Bombus terrestris colony. Behav Ecol Sociobiol 9:111119.
doi:10.1007/BF00293582
Visscher PK, Dukas R (1995) Honey bees recognize development of
nestmates ovaries. Anim Behav 49:542544. doi:10.1006
/anbe.1995.0074
Wang Y, Kaftanoglu O, Siegel AJ, Page RE Jr, Amdam GV (2010)
Surgically increased ovarian mass in the honey bee confirms link
between reproductive physiology and worker behavior. J Insect
Physiol 56:18161824. doi:10.1016/j.jinsphys.2010.07.013
West-Eberhard MJ (1996) Wasp societies as microcosms for the study of
development and evolution. Oxford University Press, Oxford
Williams JB, Roberts SP, Elekonich MM (2008) Age and natural
metabolically-intensive behavior affect oxidative stress and antiox-
idant mechanisms. Exp Gerontol 43:538549. doi:10.1016/j.
exger.2008.02.001
Worden BD, Skemp AK, Papaj DR (2005) Learning in two contexts: the
effects of interference and body size in bumblebees. J Exp Biol 208:
20452053. doi:10.1242/jeb.01582
Behav Ecol Sociobiol
... These were setup in a room with natural light, to facilitate the development of normal foraging behaviour. This sub-colony design controls for strong colony effects in this host-parasite system (Schmid-Hempel et al. 1999) 76 and whilst nest workers from queenless colonies perform better in olfactory learning tasks than those from queenright colonies (Evans et al., 2016) 77 , this factor was shared by all the sub-colonies. ...
... These were setup in a room with natural light, to facilitate the development of normal foraging behaviour. This sub-colony design controls for strong colony effects in this host-parasite system (Schmid-Hempel et al. 1999) 76 and whilst nest workers from queenless colonies perform better in olfactory learning tasks than those from queenright colonies (Evans et al., 2016) 77 , this factor was shared by all the sub-colonies. ...
Article
Full-text available
Bees are important pollinators in wild and agricultural ecosystems, and understanding the factors driving their global declines is key to maintaining these pollination services. Learning, which has been a focus of previous ecotoxicological studies in bees, may play a key role in driving colony fitness. Here we move beyond the standard single-stressor approach to ask how multiple stressors, an agrochemical (sulfoxaflor, a relatively new insecticide) and a parasite ( Crithidia bombi , a prevalent gut parasite of bumblebees), impact learning in the bumblebee Bombus terrestris . We developed a modified version of the classic proboscis extension reflex assay to assess the combined effects of acute oral sulfoxaflor exposure and infection by C. bombi on olfactory learning of bumblebee workers. We found no evidence that either sulfoxaflor, C. bombi , or their combination had any significant effect on bumblebee olfactory learning, despite their known negative impacts on other aspects of bumblebee health. This suggests that losses in cognitive ability, as measured here, are unlikely to explain the impacts of sulfoxaflor and its interactions with other stressors on bumblebees. Our novel methodology provides a model system within which to test interactive effects of other key stressors on bee health.
... Olfactory learning performance was assessed using a PER absolute conditioning paradigm (Riveros and Gronenberg, 2009;Evans et al., 2016). Identified foragers were caught and chilled on ice until they became quiescent and then harnessed within plastic syringe tubes ( Figure 1B). ...
... During each trial a bee was exposed to 5 s of unscented air then 10 s of odour-containing air. The bee was presented with 0.8 µl of sucrose solution (using a Gilmont syringe) after approximately 6 s of exposure to the odour-containing air (Evans et al., 2016). ...
Article
Full-text available
Individual animals allowed the opportunity to learn generally outperform those prevented from learning, yet, within a species the capacity for learning varies markedly. The evolutionary processes that maintain this variation in learning ability are not yet well understood. Several studies demonstrate links between fitness traits and visual learning, but the selection pressures operating on cognitive traits are likely influenced by multiple sensory modalities. In addition to vision, most animals will use a combination of hearing, olfaction (smell), gustation (taste), and touch to gain information about their environment. Some animals demonstrate individual preference for, or enhanced learning performance using certain senses in relation to particular aspects of their behaviour (e.g., foraging), whereas conspecific individuals may show different preferences. By assessing fitness traits in relation to different sensory modalities we will strengthen our understanding of factors driving observed variation in learning ability. We assessed the relationship between the olfactory learning ability of bumble bees (Bombus terrestris) and their foraging performance in their natural environment. We found that bees which failed to learn this odour-reward association had shorter foraging careers; foraging for fewer days and thus provisioning their colonies with fewer resources. This was not due to a reduced propensity to forage, but may have been due to a reduced ability to return to their colony. When comparing among only individuals that did learn, we found that the rate at which floral resources were collected was similar, regardless of how they performed in the olfactory learning task. Our results demonstrate that an ability to learn olfactory cues can have a positive impact of the foraging performance of B. terrestris in a natural environment, but echo findings of earlier studies on visual learning, which suggest that enhanced learning is not necessarily beneficial for bee foragers provisioning their colony.
... Given this, one might expect foragers to have increased cognitive function, and indeed in both ants and honeybees, individuals that forage have been shown to perform better at learning tasks than individuals based in the nest 38,39 . Similar patterns have been observed in B. terrestris colonies when the queen is present, however, in contrast to our experiment, these patterns were no longer observed upon removal of the queen 40 . In the bumblebee Bombus occidentalis, bees with more foraging experience were found to be better learners, and the foraging activity caused an increase in the size of the mushroom bodies, an area of the brain associated with learning and memory 41,42 . ...
Article
Full-text available
Supplemental feeding of wildlife populations can locally increase the density of individuals, which may in turn impact disease dynamics. Flower strips are a widely used intervention in intensive agricultural systems to nutritionally support pollinators such as bees. Using a controlled experimental semi-field design, we asked how density impacts transmission of a virus and a trypanosome parasite in bumblebees. We manipulated bumblebee density by using different numbers of colonies within the same area of floral resource. In high-density compartments, slow bee paralysis virus was transmitted more quickly, resulting in higher prevalence and level of infection in bumblebee hosts. By contrast, there was no impact of density on the transmission of the trypanosome Crithidia bombi, which may reflect the ease with which this parasite is transmitted. These results suggest that agri-environment schemes such as flower strips, which are known to enhance the nutrition and survival of bumblebees, may also have negative impacts on pollinators through enhanced disease transmission. Future studies should assess how changing the design of these schemes could minimize disease transmission and thus maximise their health benefits to wild pollinators.
... For instance, honeybees exposed to queen mandibular pheromones or brood pheromones show increased responsiveness to sucrose Page, 2000, 2001). In bumble bees, workers in colonies deprived of a queen show increased ovarian development and a higher sucrose responsiveness than broodtenders from colonies with a queen (Evans et al., 2016). In halictine bees, individuals have a reduced response threshold for excavation after they experience the presence of nestmates (Jeanson et al., 2008). ...
Article
Division of labour, whereby individuals divide the workload in a group, is a recurrent property of social living. The current conceptual framework for division of labour in social insects is provided by the response-threshold model. This model posits that the differences between individuals (i.e. between-individual variability) in responsiveness to task-associated stimuli is a key feature for task specialisation. The consistency of individual behaviours (i.e. within-individual variability) in task performance represents an additional but little-considered component driving robust patterns of division of labour. On the one hand, the presence of workers with a high level of within-individual variability presumably allows colonies to rapidly adapt to external fluctuations. On the other hand, a reduced degree of within-individual variability promotes a stricter specialisation in task performance, thereby limiting the costs of task switching. The ideal balance between flexibility and canalisation probably varies depending on the developmental stage of the colony to satisfy its changing needs. Here, I introduce the main sources of within-individual variability in behaviours in social insects and I review neural correlates accompanying the changes in behavioural flexibility. I propose the hypothesis that the positive scaling between group size and the intensity of task specialisation, a relationship consistently reported both within and between taxa, may rely on reduced within-individual variability via self-organised processes linked to the quality of brood care. Overall, I emphasise the need for a more comprehensive characterisation of the response dynamics of individuals to better understand the mechanisms shaping division of labour in social insects.
... Queen loss and the take-over of reproduction by workers affects both individual worker physiology and the structure of the whole society. On an individual level, queen loss affects brain dopamine titers (Harris and Woodring, 1995;Cuvillier-Hot and Lenoir, 2006;Shimoji et al., 2017), learning capability (Evans et al., 2016), and life span (Dixon et al., 2014;Hartmann and Heinze, 2003;Tsuji et al., 1996). On a societal level, queen loss may weaken nestmate discrimination, which in turn facilitates colony usurpation by alien conspecific queens, or by social parasites (Buschinger, 2009;Chapman et al., 2009;Tschinkel, 1996;Van Oystaeyen et al., 2013). ...
Article
Waste disposal is important for maintaining the health of animal societies. Adults and off; spring produce large amounts of waste and feces that could contain pathogens or toxins and may need to be stored away from the young or adult individuals. In social insects, the worker caste is responsible for nest maintenance, including sanitary behavior, and waste disposal strategies vary between species. However, individual task allocation is generally affected by queen presence and worker efficiency often decreases in the absence of a queen. Here we show that most (74%)colonies of the cavity-dwelling ant Temnothorax crassispinus construct up to two localized indoor ‘latrines’, which are used for defecation and only very rarely also as waste dumps. Restriction of defecation to designated areas affects the growth of mold inside the nest. Defecation strategies of colonies are furthermore affected by queen presence, with workers from queenless nests more frequently defecating outside the nest and forming latrines. As colonies do not actively avoid moldy nests, mold seems to not necessarily be a threat to the colony. While solid waste management has been more extensively studied in social insects, this study contributes a rare insight into the organization of non-easily transportable fecal waste.
... However, one exception in the literature is that ovary development increased with worker body size and these workers were often found not foraging (Smith, Graystock, & Hughes, 2016); this pattern has also been observed in honeybee workers (Roth et al., 2014). Thus, our finding of decreased trip number for largest workers might be a product of an energetic trade-off towards investment in reproductive anatomy/physiology (Roth et al., 2014;Schmid-Hempel, 1990;Smith et al., 2016) or potentially an emergent property of lower exposure of foragers to queen pheromones (Evans, Raine, & Leadbeater, 2016). ...
Article
Size‐number trade‐offs in reproduction are commonly observed in nature. Bumblebee (Bombus spp.) colonies produce workers that vary considerably in size. This variation suggests that colonies face potential size‐number trade‐offs when producing workers. Here, we estimated size‐based vital rates of Bombus vosnesenskii workers using colonies reared from wild‐caught queens. We conducted a mark–recapture study to estimate worker survival as a function of body size. We also collected data on pollen and nectar loads as well as foraging trips using a radiofrequency identification system to estimate daily resource return as a function of body size. We integrated survival and daily resource return to estimate lifetime resource collection and offset these estimates by the size‐based worker production costs. We found size‐based trade‐offs among workers of different sizes. Smaller workers had higher survival, but larger workers returned with more resources per day. The largest workers made slightly fewer foraging trips per day. Overall, larger workers made the greatest lifetime contribution to both nectar and pollen collection. However, once the benefits of larger workers are offset by their higher production costs, intermediate‐sized workers were the optimal for net resource contribution according to our models. Many previous studies have found that larger workers outperformed smaller workers with foraging and in‐nest tasks, yet these studies have not integrated multiple fitness components or worker production costs to quantify net resource contribution towards colony growth. Accounting for trade‐offs between costs and performance changed our conclusions about optimal body size from being large to being near the observed average. Similar approaches of integrating multiple vital rates may resolve apparently suboptimal life histories in other taxa. A plain language summary is available for this article. Plain Language Summary
... Given this, one might expect foragers to have increased cognitive function, and indeed in both ants and honeybees, individuals that forage have been shown to perform better at learning tasks than individuals based in the nest 38,39 . Similar patterns have been observed in B. terrestris colonies when the queen is present, however, in contrast to our experiment, these patterns were no longer observed upon removal of the queen 40 . In the bumblebee Bombus occidentalis, bees with more foraging experience were found to be better learners, and the foraging activity caused an increase in the size of the mushroom bodies, an area of the brain associated with learning and memory 41,42 . ...
Article
Full-text available
Parasites can induce behavioural changes in their host organisms. Several parasite species are known to infect bumblebees, an important group of pollinators. Task allocation within bumblebee colonies can also cause differences in behaviour. Thus, task allocation may lead to context-dependent impacts of parasites on host behaviour. This study uses Bombus terrestris and its gut trypanosome Crithidia bombi, to investigate the effects of parasitism, task allocation (foraging or nest-work) and their interactions, on olfactory learning. Prior to undergoing the olfactory learning task, bees were orally infected with a field-realistic dose of C. bombi, and observed to determine task allocation. Parasitism did not significantly affect olfactory learning, but task allocation did, with foragers being significantly more likely to learn than nest bees. There was no significant interaction between parasitism and task. These results suggest that C. bombi is unlikely to affect pollination services via changes in olfactory learning of its host if bees are under no environmental or nutritional stress. However, wild and commercial colonies are likely to face such stressors. Future studies in the field are needed to extrapolate our results to real world effects.
... Queen loss and the take-over of reproduction by workers affects both individual worker physiology and the structure of the whole society. On an individual level, queen loss affects brain dopamine titers (Harris and Woodring, 1995;Cuvillier-Hot and Lenoir, 2006;Shimoji et al., 2017), learning capability (Evans et al., 2016), and life span (Dixon et al., 2014;Hartmann and Heinze, 2003;Tsuji et al., 1996). On a societal level, queen loss may weaken nestmate discrimination, which in turn facilitates colony usurpation by alien conspecific queens, or by social parasites (Buschinger, 2009;Chapman et al., 2009;Tschinkel, 1996;Van Oystaeyen et al., 2013). ...
Article
Reproductive division of labor is a defining feature of all eusocial organisms. In the eusocial insects, non-reproductive labor is often also divided into safe tasks within the nest, such as nursing, and risky extranidal tasks, such as scouting and foraging. Within-nest (intranidal) tasks are usually performed by younger workers while older workers forage outside. This temporal specialization can be delayed, accelerated or even reversed if the need arises. Here, we studied reversion in the clonal ant Platythyrea punctata. Rather than focusing on traditional markers of temporal polyethism, such as nursing vs. foraging, we examined neglected behaviors, such as waste management and defecation, as well as the proportion of ants outside the nest and the onset of egg laying. We formed single-age groups of young intranidal, old intranidal and forager workers, and found strong differences in behavior between these groups. Intranidal workers rapidly created well-formed waste dumps (middens) and defecated within the nest, initiated egg laying sooner, and remained mostly inside the nest. Forager workers showed the reverse pattern. However, the behavior of the groups gradually converged, and after 8 weeks midden formation was similar and the proportion of ants outside the nest was identical across all groups. This study demonstrates reversible temporal polyethism in defecation and waste management, which correlates well with more traditionally studied age-related traits. Different traits seem to revert at different rates. Waste management behavior may provide an easy and low-impact way of examining behavioral reversions and allows insights into aspects of colony life not usually examined.
... Thus they relied on the assumption that the learning performance of the tested workers reflected that of workers in the field. Previous research has demonstrated the learning performance of individual bumble bees within a colony can vary significantly depending upon colony developmental stage, and worker reproductive status, at the time the assessments take place 32,33 . Consequently the mean learning performance of the laboratory-tested workers in Raine and Chittka's 19 study may not be the best representation of colony learning performance when transferred into the field. ...
Article
Full-text available
Despite widespread interest in the potential adaptive value of individual differences in cognition, few studies have attempted to address the question of how variation in learning and memory impacts their performance in natural environments. Using a novel split-colony experimental design we evaluated visual learning performance of foraging naïve bumble bees (Bombus terrestris) in an ecologically relevant associative learning task under controlled laboratory conditions, before monitoring the lifetime foraging performance of the same individual bees in the field. We found appreciable variation among the 85 workers tested in both their learning and foraging performance, which was not predicted by colony membership. However, rather than finding that foragers benefited from enhanced learning performance, we found that fast and slow learners collected food at comparable rates and completed a similar number of foraging bouts per day in the field. Furthermore, bees with better learning abilities foraged for fewer days; suggesting a cost of enhanced learning performance in the wild. As a result, slower learning individuals collected more resources for their colony over the course of their foraging career. These results demonstrate that enhanced cognitive traits are not necessarily beneficial to the foraging performance of individuals or colonies in all environments.
Article
Species' cognitive traits are shaped by their ecology, and even within a species, cognition can reflect the behavioural requirements of individuals with different roles. Social insects have a number of discrete roles (castes) within a colony and thus offer a useful system to determine how ecological requirements shape cognition. Bumblebee queens are a critical point in the lifecycle of their colony, since its future success is reliant on a single individual's ability to learn about floral stimuli while finding a suitable nest site; thus, one might expect particularly adept learning capabilities at this stage. I compared wild Bombus vosnesenskii queens and workers on their ability to learn a colour association and found that queens performed better than workers. In addition, queens of another species, B. insularis, a cuckoo species with a different lifecycle but similar requirements at this stage, performed equally well as the non-parasitic queens. To control for differences in foraging experience, I then repeated this comparison with laboratory-based B. impatiens and found that unmated queens performed better than workers. These results add to the body of work on how ecology shapes cognition and opens the door to further research in comparative cognition using wild bees.
Article
Full-text available
MULLER H & CHITTKA L (Biol Chem Sci, Queen Mary University of London, London, E1 4NS) Consistent Interindividual Differences in Discrimination Performance by Bumblebees in Colour, Shape and Odour Learning Tasks (Hymenoptera: Apidae: Bombus terrestris). - Entomol Gener 34(1/2): 001-008; Stuttgart 2012-09. - [Article] The ability to discriminate between flower-related stimuli is essential to bumblebees foraging in a natural environment, given the variety of flower features and profitability. Using differential conditioning tasks in a foraging context, it turns out that some individual bumblebees (Bombus terrestris Linnaeus 1758) were consistently better than others at discriminating rewarding from punishing stimuli across modalities (visual or olfactory) or dimension (colour or shape) of the cues used. The size of the bees did not have a significant effect on the individuals' discrimination abilities but there was a significant difference between the two colonies tested. Bees therefore appeared to differ in their ability to discriminate between stimuli both individually and at the colony level. The here presented results have implications for the study of insect cognition as well as for the study of bees' foraging behaviour and task specialisation in social insects.
Technical Report
Full-text available
Description Fit linear and generalized linear mixed-effects models. The models and their components are represented using S4 classes and methods. The core computational algorithms are implemented using the 'Eigen' C++ library for numerical linear algebra and 'RcppEigen' ``glue''.
Article
Full-text available
Several insect pheromones are multifunctional and have both releaser and primer effects. In honey bees (Apis mellifera), the queen mandibular pheromone (QMP) and e-beta-ocimene (eβ), emitted by young worker larvae, have such dual effects. There is increasing evidence that these multifunctional pheromones profoundly shape honey bee colony dynamics by influencing cooperative brood care, a fundamental aspect of eusocial insect behavior. Both QMP and eβ have been shown to affect worker physiology and behavior, but it has not yet been determined if these two key pheromones have interactive effects on hypopharyngeal gland (HPG) development, actively used in caring of larvae, and ovary activation, a component of worker reproductive physiology. Experimental results demonstrate that both QMP and eβ significantly suppress ovary activation compared to controls but that the larval pheromone is more effective than QMP. The underlying reproductive anatomy (total ovarioles) of workers influenced HPG development and ovary activation, so that worker bees with more ovarioles were less responsive to suppression of ovary activation by QMP. These bees were more likely to develop their HPG and have activated ovaries in the presence of eβ, providing additional links between nursing and reproductive physiology in support of the reproductive ground plan hypothesis. Electronic supplementary material The online version of this article (doi:10.1007/s00265-014-1811-y) contains supplementary material, which is available to authorized users.
Article
Full-text available
Animals use cues from a range of sensory modalities to discriminate stimuli and as predictors of reward. Whilst there is appreciable variation in the cognitive performance of animals, we know surprisingly little about the extent to which learning varies among individuals across different sensory modalities. Do individuals that are good at learning in one sensory modality also perform well in another (performance is correlated between modalities), or do individuals demonstrate specialisation in learning performance in one modality (trading-off performance between modalities)? We tested these hypotheses by examining the performance of 76 Bombus terrestris workers, from four colonies, in both an odour-and visual learning task. Olfactory learning was assessed using proboscis extension reflex (PER) conditioning and visual (colour) learning was examined using a well-established free-flying paradigm. Our results showed neither a correlation, nor a trade-off, in individual performance for learning tasks using different sensory modalities. However, there was considerable variation among workers within each colony in their performance in both learning tasks. This extent of interindividual variation in learning ability across sensory modalities could be adaptive for colonies dealing with changeable foraging conditions. There was also significant intercolony variation in final task performance level in the olfactory learning task, and both the strength and persistence of blue preference in the colour learning task. This is the first study to demonstrate variation in olfactory learning performance across multiple bumblebee colonies using PER conditioning, suggesting this is an effective paradigm for assessing associative olfactory learning performance both within and among colonies.
Article
Full-text available
If the cognitive performance of animals reflects their particular ecological requirements, how can we explain appreciable variation in learning ability amongst closely related individuals (e.g. foraging workers within a bumble bee colony)? One possibility is that apparent 'errors' in a learning task actually represent an alternative foraging strategy. In this study we investigate the potential relationship between foraging 'errors' and foraging success among bumble bee (Bombus terrestris) workers. Individual foragers were trained to choose yellow, rewarded flowers and ignore blue, unrewarded flowers. We recorded the number of errors (visits to unrewarded flowers) each bee made during training, then tested them to determine how quickly they discovered a more profitable food source (either familiar blue flowers, or novel green flowers). We found that error prone bees discovered the novel food source significantly faster than accurate bees. Furthermore, we demonstrate that the time taken to discover the novel, more profitable, food source is positively correlated with foraging success. These results suggest that foraging errors are part of an 'exploration' foraging strategy, which could be advantageous in changeable foraging environments. This could explain the observed variation in learning performance amongst foragers within social insect colonies.
Article
Bees derived from artificially selected high-and low-pollen-hoarding strains were tested for their proboscis extension reflex response to water and varying sucrose concentrations. High-strain bees had a lower response threshold to sucrose than low-strain bees among pre-foragers, foragers, queens and drones. Preforaging low-strain workers showed ontogenetic changes in their response threshold to sucrose which was inversely related to age. High-strain foragers were more likely to return with loads of water compared to low-strain foragers. Whereas low-strain foragers were more likely to return with loads of nectar. Low-strain nectar foragers collected nectar with significantly higher sucrose concentrations than did the high-strain nectar foragers. Alternatively, low-strain foragers were more likely to return empty compared to high-strain foragers. These studies demonstrate how a genotypically varied sensory-physiological process, the perception of sucrose, are associated with a division of labor for foraging.