ANRV330-EN53-08 ARI 2 November 2007 16:17

Evolutionary Biology
of Insect Learning
Reuven Dukas
Animal Behavior Group, Department of Psychology, Neuroscience & Behavior,
McMaster University, Hamilton, Ontario, L8S 4K1, Canada;
email: [email protected]
Annu. Rev. Entomol. 2008. 53:14560
First published online as a Review in Advance on
September 5, 2007
The Annual Review of Entomology is online at
ento.annualreviews.org
This articles doi:
10.1146/annurev.ento.53.103106.093343
Copyright c 2008 by Annual Reviews.
All rights reserved
0066-4170/08/0107-0145$20.00
Key Words
ecology, evolution, memory, fruit ies, bees, grasshoppers
Abstract
Learning and memory, dened as the acquisition and retention of
neuronal representations of new information, are ubiquitous among
insects. Recent research indicates that a variety of insects rely ex-
tensively on learning for all major life activities including feeding,
predator avoidance, aggression, social interactions, andsexual behav-
ior. There is good evidence that individuals within an insect species
exhibit genetically based variation in learning abilities and indirect
evidence linking insect learning to tness. Although insects rely on
innate behavior to successfully manage many types of variation and
unpredictability, learning may be superior to innate behavior when
dealing with features unique to time, place, or individuals. Among
insects, social learning, which can promote the rapid spread of novel
behaviors, is currently known only from a few well-studied exam-
ples insocial Hymenoptera. The prevalence and importance of social
learning in insects are still unknown. Similarly, we know little about
ecological factors that may have promoted enhanced learning abili-
ties in insects, and whether learning has signicantly contributed to
speciation in insects.
145
A
n
n
u
.

R
e
v
.

E
n
t
o
m
o
l
.

2
0
0
8
.
5
3
:
1
4
5
-
1
6
0
.

D
o
w
n
l
o
a
d
e
d

f
r
o
m

a
r
j
o
u
r
n
a
l
s
.
a
n
n
u
a
l
r
e
v
i
e
w
s
.
o
r
g
b
y

C
O
R
N
E
L
L

U
N
I
V
E
R
S
I
T
Y

o
n

0
2
/
0
8
/
0
8
.

F
o
r

p
e
r
s
o
n
a
l

u
s
e

o
n
l
y
.
ANRV330-EN53-08 ARI 2 November 2007 16:17
Learning: the
acquisition of
neuronal
representations of
new information
Social learning: the
acquisition of new
information from
other individuals
Memory: the
retention of newly
acquired information
over time
INTRODUCTION
Our understanding of insect behavior has
gradually transformed during the twentieth
century. Beginning with a widespread belief
that insect behavior is guided primarily by in-
stincts (46), researchers have accumulated ev-
idence of learning in a variety of insect species
(5). A few taxa, most notably, honey bees
(Apis mellifera) (77, 78) and parasitoid wasps
(112), were established as model systems for
research on insect learning. Furthermore, the
controversy (58) regarding associative learn-
ing in fruit ies (Drosophila melanogaster) was
resolved, and fruit ies have become a leading
model systemin work on the neurogenetics of
learning (25, 89). Consequently, at the end of
the twentieth century it was widely acknowl-
edged that learning plays an integral role in a
variety of decisions made by many insect taxa
(86).
This reviewconsiders examples pertaining
to the evolutionary biology of insect learning
and focuses on examples published since in-
sect learning was last reviewed in the Annual
Review of Entomology (87). It (a) denes learn-
ing and discusses the type of empirical data
necessary for establishing its presence or ab-
sence in a given species, (b) reviews data on
genetic variation in learning ability among in-
dividuals withininsect species, (c) discusses the
adaptive signicance of learning in insects and
details a few well-studied examples, (d ) exam-
ines the importance of social learning in in-
sects, and (e) concludes with listing promising
future directions inthe eld of insect learning.
LEARNING: DEFINITION AND
CRITICAL TESTS
Learning is the acquisition of neuronal rep-
resentations of new information. Examples
of learning include the acquisition of neu-
ronal representations of new (a) spatial envi-
ronmental congurations; (b) sensory infor-
mation such as visual, auditory, or olfactory
features; (c) associations between perceived
stimuli and environmental states; and (d ) mo-
tor patterns, for example, the sequence of
body movements involved in manipulating
a novel food. Habituation and sensitization,
typically considered simple forms of learning,
are not discussed in this review. Three basic
attributes of learning vary widely between and
within species: the ability to learn a given task,
the rate of learning a task, and the asymptote,
or the best performance achieved after exten-
sive practice. Discussions of learning implic-
itly assume the existence of memory, which is
the capacity to retain the newly acquired in-
formation for at least a short period (short-
term memory) but often over long periods
(long-term memory) (29). This review does
not discuss the distinct mechanisms associated
with learning and memory.
Whereas learning involves neuronal mod-
ication, it can be assessed only indirectly
through its potential effect on behavior. That
is, no direct method to quantify learning cur-
rently exists. This complicates learning re-
search because a series of carefully controlled
experiments are needed to infer the presence
or absence of learningwhile rulingout feasible
alternatives. For example, experiments ideally
should be conducted with observers blind to
treatments inorder toavoidobserver bias (96),
and controls must be included when relevant
to verify that either the presentation of stimuli
or environmental states alone do not generate
behavioral biases interpreted as learning (5).
Claims for lack of learning in a certain species
may also be problematic because they might
merely reect low motivation of the subjects
or behavioral deciency caused by the experi-
mental settings rather than a genuine inability
to learn. Hence the protocol must also include
a control, whichtests for some relevant behav-
ioral response by the same subjects. For exam-
ple, if one predicts differences in spatial learn-
ing between two closely related species, it is
crucial to also test subjects within the same ex-
periment for learning of colors if one predicts
no species differences in that domain (85).
The difculty of quantifying learning im-
plies that we must be cautious when inter-
preting the published literature. For example,
146 Dukas
A
n
n
u
.

R
e
v
.

E
n
t
o
m
o
l
.

2
0
0
8
.
5
3
:
1
4
5
-
1
6
0
.

D
o
w
n
l
o
a
d
e
d

f
r
o
m

a
r
j
o
u
r
n
a
l
s
.
a
n
n
u
a
l
r
e
v
i
e
w
s
.
o
r
g
b
y

C
O
R
N
E
L
L

U
N
I
V
E
R
S
I
T
Y

o
n

0
2
/
0
8
/
0
8
.

F
o
r

p
e
r
s
o
n
a
l

u
s
e

o
n
l
y
.
ANRV330-EN53-08 ARI 2 November 2007 16:17
do adult fruit ies maintain memories of ex-
periences learned in the larval stage as Tully
et al. (111) claimed, or not, as Barron &
Corbet (10) suggested? Documenting learn-
ing in fruit ies requires the use of meticulous
techniques and extensive experience. Investi-
gators in many laboratories have repeatedly
succeeded in documenting statistically signif-
icant learning scores in these insects. But ei-
ther a failure to detect learning when it occurs
or false-positive results wrongly indicating
learningcouldreect weaknesses inthe proto-
col. Hence the question of memory through
metamorphosis in fruit ies requires further
evaluation.
GENETIC VARIATION IN
LEARNING ABILITY
Learning canbe subjected to evolutionby nat-
ural selectiononly if individuals exhibit herita-
ble variation in learning that is associated with
variationintness. This sectionfocuses onge-
netically based individual variation in learning
among insects, and the following section ad-
dresses the association between learning and
tness in insects. Testing for heritable individ-
ual variation in learning has been conducted
througharticial selectionfor increasedor de-
creased learning ability. McGuire & Hirsch
(76) used the proboscis extension response
to sucrose to condition blow ies (Phormia
regina) to either saline or water. Bidirectional
selection produced bright and dull lines sig-
nicantly distinct from an unselected control
line. Holliday & Hirsch (58) later questioned
whether the blow y data from their own
laboratory reected true associative learning.
Nevertheless, similar bidirectional selection
in fruit ies employing an improved proto-
col produced divergent lines, with the propor-
tion of good learners changing from a base-
line of 19% up to 77% and down to 2% in
the bright and dull lines, respectively, after 25
generations (71). Mery & Kawecki (79) used
another protocol to select for improved learn-
ing in fruit ies. Their data indicated that,
compared with the unselected, control lines,
ies in the bright lines exhibited higher learn-
ing rates and lower rates of memory decay.
Finally, Brandes and colleagues (15, 16) doc-
umented large genetic variation in learning
in Cape honey bees (Apis mellifera capensis)
and estimated the narrow sense heritability
for learning to be around 0.4. Genetic vari-
ation in learning in honey bees (Apis mellifera)
has also been studied by Smith and colleagues
(21).
In sum, as it is with most other organis-
mal traits, there is evidence of considerable
genetic variation for both the rate of learning
and the rate of forgetting in insects. Neuro-
genetic work has identied many of the genes
involved in learning and memory in fruit ies
(25, 27). There is good evidence, however,
that many of the learning genes have broad
pleiotropic effects (18, 62). Hence, we must
take a systems-level approach to understand
possible constraints, costs, and trade-offs in-
volved in the evolution of various learning and
memory abilities.
ADAPTIVE SIGNIFICANCE
OF LEARNING
The other key condition for the evolution of
learning abilities is that genetically based indi-
vidual variationina givenlearningability is as-
sociated with tness. Because a critical evalua-
tion of this condition has not been conducted
for any animal, the following topics concern-
ing the adaptive signicance of learning are
discussed in this section: (a) when learning is
more benecial to insects than innate behav-
ior, (b) theoretical arguments and empirical
evidence of the cost of learning, (c) why insects
should learn in spite of their small body size
and short life span, and (d ) a few well-studied
examples illustrating the adaptive signicance
of learning.
Benets of Learning over Innate
Behavior
It is commonly believed that learning is an
adaptation for coping with environmental
www.annualreviews.org Evolutionary Biology of Insect Learning 147
A
n
n
u
.

R
e
v
.

E
n
t
o
m
o
l
.

2
0
0
8
.
5
3
:
1
4
5
-
1
6
0
.

D
o
w
n
l
o
a
d
e
d

f
r
o
m

a
r
j
o
u
r
n
a
l
s
.
a
n
n
u
a
l
r
e
v
i
e
w
s
.
o
r
g
b
y

C
O
R
N
E
L
L

U
N
I
V
E
R
S
I
T
Y

o
n

0
2
/
0
8
/
0
8
.

F
o
r

p
e
r
s
o
n
a
l

u
s
e

o
n
l
y
.
ANRV330-EN53-08 ARI 2 November 2007 16:17
unpredictability (3). However, all organisms,
including species that cannot learn, can read-
ily respond to a variety of factors that vary un-
predictably in space and time. For example,
Escherichia coli bacteria possess chemorecep-
tors that enable them to sense a large num-
ber of food substances and noxious chemi-
cals in their surrounding environment. They
can then employ a sophisticated system of
information processing and behavioral ma-
chinery to move toward food and away from
hazard (44, 61). Similarly, in fruit ies, fe-
males y toward the aggregation pheromone
deposited at a food source by early-arriving
females (118), and males initiate a rigid se-
quence of courtship activity in response to
conspecic females emitting sex pheromones
(56, 72, 102). Such ubiquitous cases of behav-
ioral plasticity with no learning are sufcient
as long as organisms possess the behavioral
routines that enable appropriate responses
to particular stimuli. These behavioral rou-
tines may evolve and be maintained only if
both the stimuli and the proper responses to
those stimuli remain similar over numerous
generations.
Many environmental features, however,
are unique to a certain time and place. The
ability of animals to learn about such features
expands the type and amount of information
they can respond to and, consequently, their
behavioral repertoire. For example, a bee can
acquire a neuronal representation of her nest
location, record the spatial location, odor, and
color of the best owers to forage on, and
learn a new motor pattern for handling these
owers. One can readily imagine how almost
any organismmay benet fromacquiringneu-
ronal representations of new information. In-
deed, learning may be a universal property of
all animals with a nervous system.
Costs of Learning
There are probably costs associated with
the development and maintenance of cellular
mechanisms that enable learning and mem-
ory (32). Evidence of such costs is currently
available only from a series of experiments
with fruit ies. In one study, Mery & Kawecki
(80) found that articial selection on learning
ability in adult ies, which increased learn-
ing scores in the selected lines, was associated
with reduced larval competitive ability under
low food availability. In another study, Mery
& Kawecki (81) exposed food-limited adult
ies to alternating substrate conditions, which
required use of learning for substrate choice
every two days. Flies from lines selected for
improved learning ability had lower egg-
laying rates than ies from unselected lines.
Finally, Mery & Kawecki (82) documented
that ies subjected to a training regime that
produced long-term memory died sooner in
the absence of food and water than did ies
subjected to control treatments. This set of
experiments suggests that, at least in ies with
articially selected enhanced learning ability,
learning incurs tness costs owing to energy
expenditure.
Should Insects Learn?
It has been repeatedly suggested that insects
should exhibit little learning for reasons such
as small body size and short life span (4, 75,
103). The rationale for rejecting these views
is discussed below.
Small brain. Although it is tempting to as-
sume that learning requires a central nervous
system of some size and sophistication, we
now know that this threshold must be low.
Evidence of associative learning has been well
replicated in the fruit y, whose brain con-
tains approximately 200,000 neurons (25, 28,
116), and in roundworms, Caenorhabditis ele-
gans, which possess approximately 302 neu-
rons (84, 124).
One might also argue that a small brain
limits the total amount of information that
an individual can learn and remember. How-
ever, there is currently no critical evidence
allowing us to evaluate the limits of long-
term memory in insects. Intriguingly, no lim-
its to long-term memory are known for any
148 Dukas
A
n
n
u
.

R
e
v
.

E
n
t
o
m
o
l
.

2
0
0
8
.
5
3
:
1
4
5
-
1
6
0
.

D
o
w
n
l
o
a
d
e
d

f
r
o
m

a
r
j
o
u
r
n
a
l
s
.
a
n
n
u
a
l
r
e
v
i
e
w
s
.
o
r
g
b
y

C
O
R
N
E
L
L

U
N
I
V
E
R
S
I
T
Y

o
n

0
2
/
0
8
/
0
8
.

F
o
r

p
e
r
s
o
n
a
l

u
s
e

o
n
l
y
.
ANRV330-EN53-08 ARI 2 November 2007 16:17
animal. In humans, it is well established that
individuals who learn more know more. That
is, new information supplements rather than
replaces existing relevant informationinlong-
term memory, a crucial condition that allows
the development of expertise (6, 30). It re-
mains to be tested whether the same phe-
nomenon exists in insect species that may gain
from learning large amounts of information.
Short life span. Learning is benecial only
if an individual has the opportunity to utilize
what it has learned. Such opportunity, how-
ever, may be available in any species regard-
less of its expectedlife spanbecause it may take
only seconds or minutes to learn and exploit
the newly acquired knowledge. Hence, a short
life span should not preclude learning (41).
Similarly, one might argue that if anindividual
performs a certain activity only once during
its lifetime, no learning related to this activ-
ity is expected. An opposing view, however, is
that if experience prior to performing this ac-
tivity can signicantly enhance tness, learn-
ing relevant to that activity would occur. This
indeed may be the case for some female in-
sects such as fruit ies in which many females
may mate only once. The females can acquire
substantial information about potential mate
quality from the numerous males who court
thembefore the females reach sexual maturity
and before the mature females choose a mate
(37, 38). Similarly, although many male fruit
ies may not mate throughout their lifetime
(7, 12), the experience they gain while court-
ing numerous females could increase their ex-
pected mating success (35, 36, 38, 40).
Although a short life span should not pre-
clude learning, it could limit the need for
learning. For example, insects with a single
yearly generation of a fewweeks, such as most
solitary bees, may avoid encountering many
environmental variables associated with sea-
sonal changes. A solitary bees activity may
even be synchronized with the blooming of
one or a few plant species (83), and it may ex-
perience only a limited range of weather con-
ditions, competitors, and predators during its
brief life span of only a few weeks.
As noted above, a thorough answer to the
question of whether insects should learn a cer-
tain task requires quantication of the tness
costs and benets of learning in insects under
natural settings. This has not yet been done.
Nevertheless, because learning seems to be a
universal property of animals with a central
nervous system, it is fair to adopt the a priori
assumption that an insect may learn some as-
pects of a given task. It is likely, however, that
there is a large variation among insect species
in specic learning abilities (see Prospects,
below).
A Few Examples
Perhaps the best way to appreciate the preva-
lence and importance of learning in insects is
to closely examine a few well-studied model
systems. I discuss the uses of learning in four
of the best-studied taxa: fruit ies, grasshop-
pers, parasitoid wasps, and the social honey
bees (A. mellifera) and bumble bees (Bombus
spp.).
Fruit ies. Much of the work on fruit y
learning has been stimulated by the oppor-
tunity to employ genetic tools available in
this classical model system for uncovering the
cellular mechanisms underlying learning (89,
110). Neurogenetic work has required the de-
velopment of reliable protocols for quantify-
ing the effects of learning on a variety of be-
haviors. Consequently, a variety of ingenious
procedures have been developed in the past
few decades, which suggest a broad reliance
on learning in fruit ies.
Although fruit y larvae possess limited
sensory- and information-processing abilities
compared with adults, they too can learn. In
the rst documentation of learning in fruit
y larvae, Aceves-Pina & Quinn (2) exposed
groups of third-instar larvae to three 30-s
pulses of one odor together with the appli-
cation of an electric shock. The larvae were
also exposed to three 30-s pulses of another
www.annualreviews.org Evolutionary Biology of Insect Learning 149
A
n
n
u
.

R
e
v
.

E
n
t
o
m
o
l
.

2
0
0
8
.
5
3
:
1
4
5
-
1
6
0
.

D
o
w
n
l
o
a
d
e
d

f
r
o
m

a
r
j
o
u
r
n
a
l
s
.
a
n
n
u
a
l
r
e
v
i
e
w
s
.
o
r
g
b
y

C
O
R
N
E
L
L

U
N
I
V
E
R
S
I
T
Y

o
n

0
2
/
0
8
/
0
8
.

F
o
r

p
e
r
s
o
n
a
l

u
s
e

o
n
l
y
.
ANRV330-EN53-08 ARI 2 November 2007 16:17
odor not associatedwithshock. The twotreat-
ments were alternated and separated by 90-s
breaks. In a subsequent choice test, the lar-
vae showed signicant avoidance of the odors
associated with shock. Further control ex-
periments employing the application of only
odorants or only shocks indicated no effects
of these nonassociative treatments on odorant
choice. Finally, larvae from learning-decient
mutant lines failed to show associative
learning (2).
Dukas (33) employed a protocol similar to
that used by Aceves-Pina & Quinn (2) to test
for associative learning of ecologically rele-
vant tasks in fruit y larvae. Groups of larvae
learned to prefer odors associated with high-
quality food and to avoid odors associated
with disturbance caused by simulated preda-
tion. The larvae, however, did not show sig-
nicant learning of odors associated with op-
timal temperature. Finally, Gerber et al. (54)
associated two illumination conditions (light
and dark) with a positive reinforcer, sugar, and
one of two negative reinforcers, quinine and
table salt. Experienced larvae, which were
tested individually, preferred the illumination
associated with sugar. In sum, fruit y larvae
are capable of associating either odors or light
conditions with the two types of environmen-
tal states most relevant to the larval stage, food
quality and danger.
Like the larvae, adult fruit ies can learn
to avoid odors associated with electric shock
(89) and to prefer odors associated with sugar
water (106). In addition, adult fruit ies can
learn to avoid light sources of distinct colors
associated with aversive states (shock or vio-
lent shaking) (48, 89) and to avoid ying to-
ward visual patterns associated with excessive
heat (70, 120). In short, adult fruit ies can
learn about odors, colors, and visual patterns
associated with either positive or negative
outcomes.
Both male and female fruit ies also learn
in the context of sexual behavior. The orig-
inal protocol for learning in the context of
courtship involved allowing males to court
recently mated, unreceptive females for one
hour. Compared with inexperienced males,
the males experienced with courting recently
mated females exhibited reduced courtship
of immobilized virgin females (99). Further
experiments indicated that males learn to
associate the failure to mate with specic
female pheromones (45). Experiments involv-
ing more naturalistic settings and no immobi-
lized test females indicated that male learning
is adaptive: Experience with courting unre-
ceptive, recently matedfemales causeda selec-
tive decline in subsequent courtship of mated
females but not virgin females, and experience
with courting unreceptive, immature females
resulted in a selective increase in subsequent
courtship of virgin females but not mated
females (36).
Because males rely on learning to restrict
courtship to classes of receptive females, such
learning can increase levels of assortative mat-
ing. Inexperienced D. melanogaster males nd
closely relatedD. simulans females as attractive
as intraspecic females. However, males that
experience courtship of and persistent rejec-
tion by D. simulans females learn to selectively
reduce courting such females while maintain-
ing regular high courtship levels toward con-
specic females (35).
Although much of the work on learning
in the context of courtship in fruit ies has
been restricted to males, females also have
ample opportunities for learning about po-
tential mates (38). Such learning can help the
females make better mate choice decisions.
When immature female fruit ies experienced
courtship only by small males, which are less
desirable mates than large males, the females
subsequently were more likely to mate with
small males compared with females that had
experienced courtship by large males (37).
Male fruit ies also appear to learn in the
context of aggression. In eld settings, males
defend small territories containing decaying
fruit andfemales. Larger males are more likely
to hold territories and mate (73). In labora-
tory trials, ghting males rapidly establish a
dominance hierarchy in which the winner re-
mains at a food cup containing a female and
150 Dukas
A
n
n
u
.

R
e
v
.

E
n
t
o
m
o
l
.

2
0
0
8
.
5
3
:
1
4
5
-
1
6
0
.

D
o
w
n
l
o
a
d
e
d

f
r
o
m

a
r
j
o
u
r
n
a
l
s
.
a
n
n
u
a
l
r
e
v
i
e
w
s
.
o
r
g
b
y

C
O
R
N
E
L
L

U
N
I
V
E
R
S
I
T
Y

o
n

0
2
/
0
8
/
0
8
.

F
o
r

p
e
r
s
o
n
a
l

u
s
e

o
n
l
y
.
ANRV330-EN53-08 ARI 2 November 2007 16:17
the loser retreats. When a loser in one match
was allowed to rest for 30 min and then either
rematched with the familiar winner from the
rst match, or placed with an unfamiliar win-
ner from another match, losers lunged signif-
icantly more often at unfamiliar winners than
at familiar winners (123). Individual recogni-
tion could be benecial for both males and fe-
males alsointhe context of courtshipandmate
choice. Published evidence to date, however,
only indicates that male fruit ies can learn
to distinguish among categories of females of
a distinct reproductive state (45). A role of
learning inindividual recognitionbasedonol-
factory cues is also known in solitary bees (11,
117), whereas involvement of learning in indi-
vidual recognition based on visual face mark-
ings seems to occur in social wasps (109).
The extensive work on fruit y learning
is highly illuminating for evolutionary ecol-
ogists, even though much of its focus has
been on neurogenetics. Beginning with ques-
tioning the fruit ys ability to learn (58), we
have gradually realized that even tiny, short-
lived ies employ learning in all four central
behavior categories: feeding, predator avoid-
ance, aggression, and sexual behavior. Fruit
ies likely do not possess exceptional learning
abilities relative to other insect taxa. Hence,
it is safe to assume that most other insects
also commonly employ learning in all cen-
tral aspects of life, and that such learning has
had broad inuences on insect ecology and
evolution.
Grasshoppers. Many insects must acquire
a proper balance of essential nutrients from
food sources that vary widely in composi-
tion while minimizing the consumption of
harmful compounds. Deviations fromoptimal
nutritional balance can signicantly decrease
individual growth rates and, subsequently, t-
ness (22). The role of learning in acquiring an
optimally balanced diet has been extensively
studied in locusts and grasshoppers (Acridi-
dae) (100). In a controlled laboratory study,
locusts (Locusta migratoria) fed for two days on
two synthetic foods, one lacking protein and
the other devoid of digestible carbohydrates.
One food was presented at the end of a green
tube and the other food was presented at the
end of a yellow tube, with food-color associa-
tions alternatingbetweensubjects. Duringthe
training stage, the locusts consumed the two
food types to maintain a balanced diet. The
training phase was followed with a depriva-
tion stage lasting four hours, during which
half the subjects were allowed to feed only
on a protein-decient diet and the other half
ate only a carbohydrate-decient diet. Then,
each locust was introduced into a test cham-
ber containing green and yellow tubes but no
food. The locusts preferred the color associ-
ated during training with the nutrient that was
lacking during the deprivation stage. That is,
the locusts associated visual cues with nutri-
tional qualities and relied on that learned in-
formation to seek the desired nutrient (91).
Locusts can also learn to decrease feeding on
nutrient-decient foods and avoid foods con-
taining harmful compounds (14, 65).
Locusts may achieve a balanced diet with-
out learning because the nutritional com-
position of the hemolymph directly affects
taste-receptor sensitivity, which in turn inu-
ences the locusts tendency to feed on cer-
tain diets (1, 100). What then is the bene-
t of learning? In an experiment examining
this question, grasshoppers (Schistocerca amer-
icana) were assigned into either a learning or a
random group. Each group received two syn-
thetic foods. One food consisted of a balanced
diet and the other food was carbohydrate de-
cient. Subjects in the learning group could
associate each of the two diets with distinct
tastes, colors, and spatial locations, whereas
subjects in the random group had the diet-
cue associations determined randomly twice
a day. Hence, the random grasshoppers could
not learn to associate the cues with nutritional
quality. The learning grasshoppers rapidly
learned to restrict their visits to the nutri-
tionally balanced food, whereas the random
grasshoppers kept visiting each food type at
equal frequencies. The random grasshoppers,
however, gradually increased the proportion
www.annualreviews.org Evolutionary Biology of Insect Learning 151
A
n
n
u
.

R
e
v
.

E
n
t
o
m
o
l
.

2
0
0
8
.
5
3
:
1
4
5
-
1
6
0
.

D
o
w
n
l
o
a
d
e
d

f
r
o
m

a
r
j
o
u
r
n
a
l
s
.
a
n
n
u
a
l
r
e
v
i
e
w
s
.
o
r
g
b
y

C
O
R
N
E
L
L

U
N
I
V
E
R
S
I
T
Y

o
n

0
2
/
0
8
/
0
8
.

F
o
r

p
e
r
s
o
n
a
l

u
s
e

o
n
l
y
.
ANRV330-EN53-08 ARI 2 November 2007 16:17
of time spent feeding on the balanced diet,
suggesting that they relied on a nonlearn-
ing mechanism such as the change in taste-
receptor sensitivity described above. Never-
theless, the overall time spent feeding on the
balanced diet was over 99% for the learn-
ing grasshoppers and only 87% for the ran-
dom grasshoppers. Furthermore, the random
grasshoppers did not feed on the balanced diet
as regularly as the learning grasshoppers did
because they approached the dish with the de-
cient diet equally as often as they approached
the dish with the balanced diet, which resulted
in aborting the meal and resting until com-
mencing another feeding attempt. The be-
havioral differences between the treatments
translated into a 20% higher growth rate in
the learning grasshoppers compared with the
randomgrasshoppers (39). The tness benet
fromlearning would also be signicant in nat-
ural settings, where learning could also help
grasshoppers minimize dangerous travel.
Parasitoid wasps. Parasitoid wasps typically
rely on innate mechanisms to identify and at-
tack their host. The host, however, may oc-
cur on a variety of substrates, which often
are much easier to locate than the host it-
self. Because there are typically positive spa-
tial and temporal correlations in host distri-
bution, parasitoid wasps can probably detect
more hosts if they learn to seek the substrate
onwhichthey have recently encounteredtheir
host. Indeed, rapid learning of cues associated
with host availability has been documented
in a variety of parasitoid wasps (112). In an
experiment with the braconid parasitoid Mi-
croplitis croceipes, females were divided into
two treatment groups, hungry and satiated.
Each of the groups was further divided into
two experience treatments in which host (lar-
vae of Helicoverpa zea) and food (sugar wa-
ter) were presented with distinct novel odors.
Half of the hungry wasps and half of the sa-
tiated wasps experienced sugar water associ-
ated with chocolate odor and the host asso-
ciated with vanilla odor. The other half of
each group experienced sugar water associ-
ated with vanilla odor and host associated with
chocolate odor. A test conducted 40 min af-
ter the training phase indicated that the sati-
ated wasps preferred the odor associated with
their host during training, whereas the hun-
gry wasps preferred the odor associated with
food during training (68). M. croceipes para-
sitoid wasps have also been shown to learn
colors, shapes, and visual patterns associated
with their host (115).
The effects of learning on locating host-
substrate were also documented in eld
settings. Females of the Drosophila larval par-
asitoid Leptopilina heterotoma that had expe-
rienced parasitizing larvae infesting decaying
mushrooms landed more often on decay-
ing mushroom baits, whereas wasps that
had experienced parasitizing larvae infest-
ing fermenting apples landed more often on
fermenting apple baits (88).
Honey bees and bumble bees. Extensive
researchonhoney bees (A. mellifera) andbum-
ble bees (Bombus spp.) has provided us withthe
best information about the uses of learning
by insects in the eld. Although this section
focuses on honey bees, I provide supplemen-
tary data on bumble bees where relevant. Be-
fore honey bees initiate their last role in life
as foragers, they carry out a few orientation
ights, each of which lasts several minutes.
These ights enable bees to learn the exact
spatial location of the hive and the geography
in the vicinity of the hive (19, 114). Whereas
it may be relatively easy to locate a human-
made hive in human-modied surroundings,
which provide prominent landmarks, learning
the spatial location of the bees natural nest, a
tree cavity in the forest, could be challenging,
even for humans.
Young foragers (observers) may rely on in-
formation from experienced foragers (mod-
els) to locate protable owers. The model
bees waggle dances, which are performed in
the dark hive, encode information about the
direction of and distance to the owers they
have visited. Observer bees also learn the o-
ral odors associated with model bees (47, 98,
152 Dukas
A
n
n
u
.

R
e
v
.

E
n
t
o
m
o
l
.

2
0
0
8
.
5
3
:
1
4
5
-
1
6
0
.

D
o
w
n
l
o
a
d
e
d

f
r
o
m

a
r
j
o
u
r
n
a
l
s
.
a
n
n
u
a
l
r
e
v
i
e
w
s
.
o
r
g
b
y

C
O
R
N
E
L
L

U
N
I
V
E
R
S
I
T
Y

o
n

0
2
/
0
8
/
0
8
.

F
o
r

p
e
r
s
o
n
a
l

u
s
e

o
n
l
y
.
ANRV330-EN53-08 ARI 2 November 2007 16:17
114). A variety of experiments (43, 55), as well
as recent observations using harmonic radar
(95), indicate that observer bees learn the di-
rection and distance information encoded in
the waggle dance and rely on that information
to arrive in the general vicinity of the owers.
The bees are then assisted by olfactory and
visual cues from the owers and perhaps also
by following model bees and pheromones (51,
95, 105, 114).
Bees visiting a certain ower patch for
the rst time exhibit characteristic orienta-
tion ights, which are used during learning
the spatial location of the patch (23, 42, 67).
Under natural settings, honey bees can learn
the spatial information necessary for foraging
more than 10 km from the hive (13, 98, 113).
Within a ower eld, honey bees learn to re-
strict visits to a small area, which they keep
revisiting over successive trips and days (50).
This site delity probably allows bees to learn
subtleties such as which individual plants of-
fer higher reward (nectar and pollen) rates. In
bumble bees, eld experiments suggest that
foragers can employ spatial learning to direct
more visits to individual plants that provide
higher nectar secretion rates (20). Controlled
laboratory studies have conrmed that bum-
ble bees can rely on spatial learning to restrict
visits to rewarding owers within a patch that
contains rewarding and nonrewarding owers
(17).
Honey bees can also learn how to bet-
ter handle certain owers (94). In controlled
studies with bumble bees, nave foragers typi-
cally located nectar on their rst visits to sim-
ple owers witheasy access tonectar. The bees
then took several minutes to reach asymptotic
handling speed. Oncomplex owers withnec-
tar hiddenina closedtube or unusual location,
more than 50% of nave bees failed to locate
nectar on their rst visit. Bees that succeeded
at locating nectar took up to one hour of for-
agingtoreachasymptotic handlingspeed(63).
Overall, learning a variety of tasks appears
to translate into a gradual increase in perfor-
mance over the lives of forager honey bees.
Bees that initiate foraging have a low food-
delivery rate, measured as the weight of o-
ral reward delivered to the hive per unit of
foraging-trip duration. Foragers then exhibit
a gradual increase in the food-delivery rate,
reaching a peak after 7 to 10 days, which is ap-
proximately the expected life span of foragers
(41, 97). Factors other than learning seem to
contribute relatively little to the observed in-
crease in foraging performance throughout a
foragers life. Enzymatic analyses of the ight
muscles of bees suggest that bees initiating
foraging are close to asymptotic physiological
performance (97). Similarly, there is no evi-
dence of an increase in foraging effort with
forager experience. In sum, extensive reliance
on social and individual learning to locate
nutrient-rich food sources, to navigate suc-
cessfully between food sources and the hive,
and to handle owers efciently make learn-
ing a dominant element in determining the
contribution of forager honey bees to colony
tness.
SOCIAL LEARNING IN INSECTS
Social learning, denedas learning fromother
individuals, is important because it allows
fast spread of novel behaviors within and be-
tween generations. Individuals may benet
more from social learning because it is faster
than individual learning and saves the tness
costs of errors associated with inexperience
(52). Social learning has been studied mostly
inmammals, birds, andsh(57). Ininsects, so-
cial learning has been unambiguously demon-
strated only in social Hymenoptera but this
probably reects limited research effort.
The most celebrated case of social learning
in an insect is the waggle dance of honey bees
discussed above. More limited cases of social
learning about food sources are also known in
other genera of the Apidae, including stingless
bees (Meliponini) and bumble bees (Bombus
spp.) (26, 114). Experienced ants (Temnotho-
rax albipennis) also employ social learning to
lead inexperienced nestmates to food sources
by using a technique known as tandem run-
ning. On average, when followed by an
www.annualreviews.org Evolutionary Biology of Insect Learning 153
A
n
n
u
.

R
e
v
.

E
n
t
o
m
o
l
.

2
0
0
8
.
5
3
:
1
4
5
-
1
6
0
.

D
o
w
n
l
o
a
d
e
d

f
r
o
m

a
r
j
o
u
r
n
a
l
s
.
a
n
n
u
a
l
r
e
v
i
e
w
s
.
o
r
g
b
y

C
O
R
N
E
L
L

U
N
I
V
E
R
S
I
T
Y

o
n

0
2
/
0
8
/
0
8
.

F
o
r

p
e
r
s
o
n
a
l

u
s
e

o
n
l
y
.
ANRV330-EN53-08 ARI 2 November 2007 16:17
Speciation: the
formation of a new
biological species
observer ant, a model ant moves fromthe nest
tothe foodsource 25%slower thanwhentrav-
eling alone. The model ant adjusts her speed,
slowing down when the observer ant is de-
layed and accelerating when the observer is
following closely. Tandem running in the ant
T. albipennis was formally described as a case
of teaching (49), which is a highly restrictive
category of social learning. Currently, only
one other case of teaching in nonhuman ani-
mals has been conrmed (107). In addition to
recruiting nestmates to distant food sources,
social bees can also copy the ower choice
of experienced foragers. Experiments in two
laboratories indicated that inexperienced ob-
server bees (Bombus spp.) were more likely to
land on the ower type visited by model bees
than on an unvisited alternative (64, 121).
PROSPECTS
This review indicates that a variety of insects
rely onlearning to enhance all major life activ-
ities, including feeding, antipredatory behav-
ior, aggression, social interactions, courtship,
and mate choice. Given that insect learning
is now well documented, future research can
focus on a variety of topics concerned with
the evolution of learning and with the effects
of learning on insect ecology and evolution.
These issues are outlined below.
Ecological Variables Associated with
Enhanced Learning Abilities
There is widespread interest in the evolu-
tion of learning, cognition, and intelligence.
To this end, work on vertebrates has suc-
cessfully identied positive correlations be-
tween specic ecological needs and either the
necessary learning abilities required to meet
these needs or the relative volumes of spe-
cic brain parts housing the cognitive traits
that process these needs (34). The striking
lack of comparable data from insects indi-
cates that researchis biasedtowardvertebrates
or that negative data have remained unpub-
lished. There is a large interspecies variation
in the relative size of brain parts such as the
mushroom body (104), but that variation has
not yet been tightly linked to insect learning
and ecology. Fruitful lines of research within
this area include testing the predictions that
(a) nesting insects that shuttle between their
nest and food sources would have better spa-
tial learning and memory than closely related
nonnesting species (31), (b) social learning and
behavioral exibility (innovation) are associ-
ated with larger volumes of relevant brain
parts such as the mushroombody (66, 93), and
(c) nectar feeders would be less likely to exhibit
taste-aversion learning than would closely re-
lated foliage feeders (90). A related promising
research program should examine the neu-
rogenetic mechanisms underlying between-
species variation in learning and memory abil-
ities (101, 122).
Effects of Learning on Speciation
in Insects
Behavior in general and learning in particular
may be major forces driving speciation (9, 74,
119). This widespread assertion, however, is
not yet broadly supported by theory or data
(34). Insects have been employed extensively
inthe study of speciation(24, 59, 69) andcould
readily be used for critical tests of the role of
learning in speciation.
Insect learning can contribute to speci-
ation in at least two ways. First, if adults
can learn to seek the substrate they fed on
as larvae, and if adults typically mate at
the substrate, the resulting assortative mat-
ing could lead to speciation (108). At least in
D. melanogaster, parts of the mushroom body,
the brain part involved in olfactory learning,
remain intact during metamorphosis, poten-
tially allowing memory to be maintained from
larvae to adults (8). A few reports have indi-
cated the transfer of memory from larvae to
adults (53, 92, 111). Furthermore, two alter-
native learning scenarios do not require mem-
ory throughmetamorphosis. First, insome in-
sects, the larvae pupate next to the substrate;
therefore eclosing adults can learn directly the
154 Dukas
A
n
n
u
.

R
e
v
.

E
n
t
o
m
o
l
.

2
0
0
8
.
5
3
:
1
4
5
-
1
6
0
.

D
o
w
n
l
o
a
d
e
d

f
r
o
m

a
r
j
o
u
r
n
a
l
s
.
a
n
n
u
a
l
r
e
v
i
e
w
s
.
o
r
g
b
y

C
O
R
N
E
L
L

U
N
I
V
E
R
S
I
T
Y

o
n

0
2
/
0
8
/
0
8
.

F
o
r

p
e
r
s
o
n
a
l

u
s
e

o
n
l
y
.
ANRV330-EN53-08 ARI 2 November 2007 16:17
larval-substrate characteristics (60). Second,
eclosing adults may smell substrate odors re-
maining on the pupae from the larval stage
(10). In short, there is solid evidence that adult
insects can learn about their larval substrate
and exhibit signicant preference for this sub-
strate over available alternatives. Further re-
search may examine the role that such learned
preference plays in speciation.
The second way in which insect learn-
ing can contribute to speciation involves mate
choice. Work on D. melanogaster reviewed
above indicates that both males and females
learn in the context of mate choice. In fe-
males, learning about the variety of locally
available males might narrow the range of
acceptable mates and hence increase assor-
tative mating (37). In males, learning de-
creases interspecic courtship (35). Learning
in the context of sexual behavior is proba-
bly prevalent in insects but current empir-
ical data are limited (38). We also have to
evaluate the importance of such learning in
speciation.
SUMMARY POINTS
1. Learning is probably a universal property of insects, which rely on learning for all
major life functions.
2. Genetically based individual variation in learning has been documented in a fewinsect
species.
3. The widespread assertion that insects may exhibit little learning owing to their small
brain and brief life span has been rejected by recent theory and data.
4. Social learning is currently known in social Hymenoptera but may be prevalent among
other insects.
5. Because occurrences of learning are well documented in a variety of insects, future
research can build on that information to examine ecological features associated with
enhanced learning abilities and how learning inuences evolution.
DISCLOSURE STATEMENT
The author is not aware of any biases that might be perceived as affecting the objectivity of
this review.
ACKNOWLEDGMENTS
I thank Bernie Roitberg and Lauren Taylor for comments on the manuscript. My work has
been supported by the Natural Sciences and Engineering Research Council of Canada.
LITERATURE CITED
1. Abisgold JD, Simpson SJ. 1988. The effect of dietary-protein levels and hemolymph
composition on the sensitivity of the maxillary palp chemoreceptors of locusts. J. Exp.
Biol. 135:21529
2. Aceves-Pina EO, Quinn WG. 1979. Learning in normal and mutant Drosophila larvae.
Science 206:9396
3. Alcock J. 2005. Animal Behavior. Sunderland, MA: Sinauer. 8th ed.
www.annualreviews.org Evolutionary Biology of Insect Learning 155
A
n
n
u
.

R
e
v
.

E
n
t
o
m
o
l
.

2
0
0
8
.
5
3
:
1
4
5
-
1
6
0
.

D
o
w
n
l
o
a
d
e
d

f
r
o
m

a
r
j
o
u
r
n
a
l
s
.
a
n
n
u
a
l
r
e
v
i
e
w
s
.
o
r
g
b
y

C
O
R
N
E
L
L

U
N
I
V
E
R
S
I
T
Y

o
n

0
2
/
0
8
/
0
8
.

F
o
r

p
e
r
s
o
n
a
l

u
s
e

o
n
l
y
.
ANRV330-EN53-08 ARI 2 November 2007 16:17
4. Alexander RD, Marshall DC, Cooley JR. 1997. Evolutionary perspectives on insect mat-
ing. In The Evolution of Mating Systems in Insects and Arachnids, ed. JC Choe, BJ Crespi,
pp. 431. New York: Cambridge Univ. Press
5. Alloway TM. 1972. Learning and memory in insects. Annu. Rev. Entomol. 17:4356
6. Anderson JR. 2000. Learning and Memory. New York: Wiley
7. Andersson M. 1994. Sexual Selection. Princeton, NJ: Princeton Univ. Press
8. Armstrong JD, de-Belle JS, Wang ZS, Kaiser K. 1998. Metamorphosis of the mushroom
bodies: large-scale rearrangements of the neural substrates for associative learning and
memory in Drosophila. Learn. Mem. 5:10214
9. Baldwin JM. 1896. A new factor in evolution. Am. Nat. 30:44151
10. Barron AB, Corbet SH. 1999. Preimaginal conditioning in Drosophila revisited. Anim.
Behav. 58:62128
11. Barrows EM, Bell WJ, Michener CD. 1975. Individual odor differences and their social
functions in insects. Proc. Natl. Acad. Sci. USA 72:282428
12. Bateman AJ. 1948. Intra-sexual selection in Drosophila. Heredity 2:34968
13. Beekman M, Ratnieks FLW. 2000. Long-range foraging by the honeybee, Apis mellifera
L. Funct. Ecol. 14:49096
14. Behmer ST, Elias DO, Bernays EA. 1999. Post-ingestive feedbacks and associative learn-
ing regulate the intake of unsuitable sterols in a generalist grasshopper. J. Exp. Biol.
202:73948
15. Brandes C. 1991. Genetic differences in learning behavior in honeybees (Apis mellifera
capensis). Behav. Genet. 21:27194
16. Brandes C, Frisch B, Menzel R. 1988. Time-course of memory formation differs in honey
bee lines selected for good and poor learning. Anim. Behav. 36:98185
17. Burns JG, Thomson JD. 2006. A test of spatial memory and movement patterns of bum-
blebees at multiple spatial and temporal scales. Behav. Ecol. 17:4855
18. Butcher LM, Kennedy JKJ, PlominR. 2006. Generalist genes and cognitive neuroscience.
Curr. Opin. Neurobiol. 16:14551
19. Capaldi EA, Smith AD, Osborne JL, Fahrbach SE, Farris SM, et al. 2000. Ontogeny of
orientation ight in the honeybee revealed by harmonic radar. Nature 403:53740
20. Cartar RV. 2004. Resource-tracking by bumble bees: responses to plant-level differences
in quality. Ecology 85:276471
21. Chandra SBC, Hosler JS, Smith BH. 2000. Heritable variation for latent inhibition and its
correlationwithreversal learning inhoneybees (Apis mellifera). J. Comp. Psychol. 114:8697
22. Chapman RF. 1998. The Insects. Cambridge, UK: Cambridge Univ. Press
23. Collett TS, Collett M. 2002. Memory use in insect visual navigation. Nat. Rev. Neurosci.
3:54252
24. Coyne JA, Orr HA. 1989. Patterns of speciation in Drosophila. Evolution 43:36281
25. Davis RL. 2005. Olfactory memory formation in Drosophila: from molecular to systems
neuroscience. Annu. Rev. Neurosci. 28:275302
26. Dornhaus A, Chittka L. 1999. Evolutionary origins of bee dances. Nature 401:38
27. Dubnau J, Chiang A-S, Tully T. 2003. Neural substrates of memory: from synapse to
system. J. Neurobiol. 54:23853
28. Dubnau J, Tully T. 1998. Gene discovery in Drosophila: new insights for learning and
memory. Annu. Rev. Neurosci. 21:40744
29. Dudai Y. 2002. Memory from A to Z: Keywords, Concepts, and Beyond. Oxford: Oxford Univ.
Press
30. Dukas R. 1998. Constraints on information processing and their effects on behavior. In
Cognitive Ecology, ed. R Dukas, pp. 89127. Chicago: Univ. Chicago Press
156 Dukas
A
n
n
u
.

R
e
v
.

E
n
t
o
m
o
l
.

2
0
0
8
.
5
3
:
1
4
5
-
1
6
0
.

D
o
w
n
l
o
a
d
e
d

f
r
o
m

a
r
j
o
u
r
n
a
l
s
.
a
n
n
u
a
l
r
e
v
i
e
w
s
.
o
r
g
b
y

C
O
R
N
E
L
L

U
N
I
V
E
R
S
I
T
Y

o
n

0
2
/
0
8
/
0
8
.

F
o
r

p
e
r
s
o
n
a
l

u
s
e

o
n
l
y
.
ANRV330-EN53-08 ARI 2 November 2007 16:17
31. Dukas R. 1998. Evolutionary ecology of learning. In Cognitive Ecology, ed. R Dukas,
pp. 12974. Chicago: Univ. Chicago Press
32. Dukas R. 1999. Costs of memory: ideas and predictions. J. Theor. Biol. 197:4150
33. Dukas R. 1999. Ecological relevance of associative learning in fruit y larvae. Behav. Ecol.
Sociobiol. 45:195200
34. Dukas R. 2004. Evolutionary biology of animal cognition. Annu. Rev. Ecol. Evol. Syst.
35:34774
35. Dukas R. 2004. Male fruit ies learn to avoid interspecic courtship. Behav. Ecol. 15:695
98
36. Dukas R. 2005. Experience improves courtship in male fruit ies. Anim. Behav. 69:12039
37. One in a series
of papers
documenting the
adaptive
signicance of
learning in the
context of sexual
behavior in insects.
37. Dukas R. 2005. Learning affects mate choice in female fruit ies. Behav. Ecol.
16:8004
38. Dukas R. 2006. Learning in the context of sexual behavior in insects. Anim. Biol. 56:125
41
39. Dukas R, Bernays EA. 2000. Learning improves growth rate in grasshoppers. Proc. Natl.
Acad. Sci. USA 97:263740
40. Dukas R, Clark CW, Abbott K. 2006. Courtship strategies of male insects: When is
learning advantageous? Anim. Behav. 72:1395404
41. Dukas R, Visscher PK. 1994. Lifetime learning by foraging honey bees. Anim. Behav.
48:100712
42. Dyer FC. 1998. Cognitive ecology of navigation. In Cognitive Ecology, ed. R Dukas,
pp. 20160. Chicago: Univ. Chicago Press
43. Dyer FC. 2002. The biology of the dance language. Annu. Rev. Entomol. 47:91749
44. Eisenbach M, Lengeler JW. 2004. Chemotaxis. London: Imperial College Press. 499 pp.
45. Ejima A, Smith BPC, Lucas C, Levine JD, Grifth LC. 2005. Sequential learning
of pheromonal cues modulates memory consolidation in trainer-specic associative
courtship conditioning. Curr. Biol. 15:194206
46. Fabre J-H, Teixeira de Mattos A, Miall B. 1918. The Wonders of Instinct. New York: The
Century Co. 322 pp.
47. Farina WM, Gruter C, Diaz PC. 2005. Social learningof oral odours inside the honeybee
hive. Proc. R. Soc. London Sci. Ser. B 272:192328
48. Folkers E. 1982. Visual learning and memory of Drosophila melanogaster wild type CS
and the mutants dunce1, amnesiac, turnip and rutabaga. J. Insect Physiol. 28:53539
49. The rst report
of teaching in a
nonhuman animal.
49. Franks NR, RichardsonT. 2006. Teaching intandem-running ants. Nature 439:153
50. Free JB. 1970. Insect Pollination of Crops. London: Academic
51. Free JB. 1987. Pheromones of Social Bees. London: Chapman and Hall
52. Galef BG, Laland KN. 2005. Social learning in animals: empirical studies and theoretical
models. Bioscience 55:48999
53. Gandol M, Mattiacci L, Dorn S. 2003. Preimaginal learning determines adult response
to chemical stimuli in a parasitic wasp. Proc. R. Soc. London B 270:262329
54. Gerber B, Scherer S, Neuser K, Michels B, Hendel T, et al. 2004. Visual learning in
individually assayed Drosophila larvae. J. Exp. Biol. 207:17988
55. Gould JL. 1975. Honey bee recruitment. Science 189:68593
56. Hall JC. 1994. The mating of a y. Science 264:170214
57. Heyes CM, Galef BG, eds. 1996. Social Learning in Animals. San Diego, CA: Academic
58. Holliday M, Hirsch J. 1986. A comment on the evidence for learning in Diptera. Behav.
Genet. 16:43947
59. Howard DJ, Berlocher SH, eds. 1998. Endless Forms: Species and Speciation. New York:
Oxford Univ. Press. 470 pp.
www.annualreviews.org Evolutionary Biology of Insect Learning 157
A
n
n
u
.

R
e
v
.

E
n
t
o
m
o
l
.

2
0
0
8
.
5
3
:
1
4
5
-
1
6
0
.

D
o
w
n
l
o
a
d
e
d

f
r
o
m

a
r
j
o
u
r
n
a
l
s
.
a
n
n
u
a
l
r
e
v
i
e
w
s
.
o
r
g
b
y

C
O
R
N
E
L
L

U
N
I
V
E
R
S
I
T
Y

o
n

0
2
/
0
8
/
0
8
.

F
o
r

p
e
r
s
o
n
a
l

u
s
e

o
n
l
y
.
ANRV330-EN53-08 ARI 2 November 2007 16:17
60. Jaenike J. 1982. Environmental modication of oviposition behavior in Drosophila. Am.
Nat. 119:784802
61. Koshland D. 1980. Bacterial Chemotaxis as a Model Behavioral System. New York: Raven
62. Kovas Y, Plomin R. 2006. Generalist genes: implications for the cognitive sciences. Trends
Cogn. Sci. 10:198203
63. Laverty TM. 1994. Bumble bee learning and ower morphology. Anim. Behav. 47:53145
64. Leadbeater E, Chittka L. 2005. A new mode of information transfer in foraging bumble-
bees? Curr. Biol. 15:R447R48
65. Lee JC, Bernays EA. 1990. Food tastes and toxic effects: associative learning by the
polyphagous grasshopper Schistocerca americana (Drury) (Orthoptera: Acrididae). Anim.
Behav. 39:16373
66. Lefebvre L, Whittle P, Lascaris E, FinkelsteinA. 1997. Feeding innovations and forebrain
size in birds. Anim. Behav. 53:54960
67. Lehrer M. 1993. Why do bees turn back and look? J. Comp. Physiol. A 172:54963
68. Lewis WL, Takasu K. 1990. Use of learned odours by a parasitic wasp in accordance with
host and food needs. Nature 348:63536
69. Linn C, Feder JL, Nojima S, Dambroski HR, Berlocher SH, Roelofs W. 2003. Fruit
odor discrimination and sympatric host race formation in Rhagoletis. Proc. Natl. Acad. Sci.
USA 100:1149093
70. Liu G, Seiler H, Wen A, Zars T, Ito K, et al. 2006. Distinct memory traces for two visual
features in the Drosophila brain. Nature 439:55156
71. Lofdahl KL, Holliday M, Hirsch J. 1992. Selection for conditionability in Drosophila
melanogaster. J. Comp. Psychol. 106:17283
72. Manoli DS, Meissner GW, Baker BS. 2006. Blueprints for behavior: genetic specication
of neural circuitry for innate behaviors. Trends Neurosci. 29:44451
73. MarkowTA. 1988. Reproductive behavior of Drosophila melanogaster and D. nigrospiracula
in the eld and in the laboratory. J. Comp. Psychol. 102:16973
74. Mayr E. 1963. Animal Species and Evolution. Cambridge, MA: Harvard Univ. Press
75. Mayr E. 1974. Behavior programs and evolutionary strategies. Am. Sci. 62:65059
76. McGuire TR, Hirsch J. 1977. Behavior genetic analysis of Formia regina: conditioning,
reliable individual differences, and selection. Proc. Natl. Acad. Sci. USA 74:519397
77. Menzel R, Erber J, Masuhr T. 1974. Learningandmemory inthe honeybee. InExperimen-
tal Analysis of Insect Behavior, ed. L Barton-Browne, pp. 195217. Berlin: Springer-Verlag
78. Menzel R, Muller U. 1996. Learning and memory in honeybees: from behavior to neural
substrates. Annu. Rev. Neurosci. 19:379404
79. Mery F, Kawecki TJ. 2002. Experimental evolution of learning ability in fruit ies. Proc.
Natl. Acad. Sci. USA 99:1427479
80. Mery F, Kawecki TJ. 2003. A tness cost of learning ability in Drosophila melanogaster.
Proc. R. Soc. London Sci. Ser. B 270:246569
81. Mery F, Kawecki TJ. 2004. An operating cost of learning in Drosophila melanogaster. Anim.
Behav. 68:58998
82. One in a series
of papers
quantifying the
cost of learning.
82. Mery F, Kawecki TJ. 2005. A cost of long-term memory in Drosophila. Science
308:1148
83. Minckley RL, Wcislo WT, Yanega D, Buchmann SL. 1994. Behavior and phenology of a
specialist bee (Dieunomia) and sunower (Helianthus) pollen availability. Ecology 75:1406
19
84. Nuttley WM, Atkinson-Leadbeater KP, van der Kooy D. 2002. Serotonin mediates food-
odor associative learning in the nematode Caenorhabditis elegans. Proc. Natl. Acad. Sci. USA
99:1244954
158 Dukas
A
n
n
u
.

R
e
v
.

E
n
t
o
m
o
l
.

2
0
0
8
.
5
3
:
1
4
5
-
1
6
0
.

D
o
w
n
l
o
a
d
e
d

f
r
o
m

a
r
j
o
u
r
n
a
l
s
.
a
n
n
u
a
l
r
e
v
i
e
w
s
.
o
r
g
b
y

C
O
R
N
E
L
L

U
N
I
V
E
R
S
I
T
Y

o
n

0
2
/
0
8
/
0
8
.

F
o
r

p
e
r
s
o
n
a
l

u
s
e

o
n
l
y
.
ANRV330-EN53-08 ARI 2 November 2007 16:17
85. Olson DJ, Kamil AC, Balda RP, Nims PJ. 1995. Performance of four seed-caching corvid
species in operant tests of nonspatial and spatial memory. J. Comp. Psychol. 109:17381
86. Papaj DR, Lewis AC. 1993. Insect Learning. New York: Chapman and Hall
87. Papaj DR, Prokopy RJ. 1989. Ecological and evolutionary aspects of learning in phy-
tophagous insects. Annu. Rev. Entomol. 34:31550
88. Papaj DR, Vet LEM. 1990. Odor learning and foraging success in the parasitoid, Lep-
topilina heterotoma. J. Chem. Ecol. 16:313749
89. QuinnWG, Harris WA, Benzer S. 1974. Conditioned behavior in Drosophila melanogaster.
Proc. Natl. Acad. Sci. USA 71:70812
90. Ratcliffe JM, Fenton MB, Galef BG. 2003. An exception to the rule: Common vampire
bats do not learn taste aversions. Anim. Behav. 65:38589
91. Raubenheimer D, Tucker D. 1997. Associative learning by locusts: pairing of visual cues
with consumption of protein and carbohydrate. Anim. Behav. 54:144959
92. Ray S. 1999. Survival of olfactory memory through metamorphosis in the y Musca
domestica. Neurosci. Lett. 259:3740
93. Reader SM, Laland KN. 2002. Social intelligence, innovation, and enhanced brain size
in primates. Proc. Natl. Acad. Sci. USA 99:443641
94. Reinhardt JF. 1952. Some responses of honey bees to alfalfa owers. Am. Nat. 86:25775
95. Riley JR, Greggers U, Smith AD, Reynolds DR, Menzel R. 2005. The ight paths of
honeybees recruited by the waggle dance. Nature 435:2057
96. Rosenthal R, Fode KL. 1963. The effect of experimenter bias on the performance of the
albino rat. Behav. Sci. 8:18389
97. Schippers M-P, Dukas R, Smith RW, Wang J, Smolen K, McClelland GB. 2006. Lifetime
performance in foraging honeybees: behaviour and physiology. J. Exp. Biol. 209:382836
98. Seeley TD. 1996. The Wisdom of the Hive. Cambridge, MA: Harvard Univ. Press
99. Siegel RW, Hall JC. 1979. Conditioned courtship in Drosophila and its mediation by
association of chemical cues. Proc. Natl. Acad. Sci. USA 76:343034
100. Simpson SJ, Raubenheimer D. 2000. The hungry locust. Adv. Study Behav. 29:144
101. Smid HM, Wang G, Bukovinszky T, Steidle JLM, Bleeker MAK, et al. 2007. Species-
specic acquisition and consolidation of long-term memory in parasitic wasps. Proc. R.
Soc. London Ser. B 274:153946
102. Spieth HT. 1974. Courtship behavior in Drosophila. Annu. Rev. Entomol. 19:383406
103. Staddon JER. 1983. Adaptive Behavior and Learning. Cambridge, MA: Cambridge Univ.
Press
104. Strausfeld NJ, Hansen L, Li YS, Gomez RS, Ito K. 1998. Evolution, discovery, and
interpretations of arthropod mushroom bodies. Learn. Mem. 5:1137
105. Tautz J, Sandeman DC. 2003. Recruitment of honeybees to nonscented food sources.
J. Comp. Physiol. A 189:293300
106. Tempel BL, Bonini N, Dawson DR, Quinn WG. 1983. Reward learning in normal and
mutant Drosophila. Proc. Natl. Acad. Sci. USA 80:148286
107. Thornton A, McAuliffe K. 2006. Teaching in wild meerkats. Science 313:22729
108. Thorpe W, Jones F. 1937. Olfactory conditioning in a parasitic insect and its relation to
the problem of host selection. Proc. R. Soc. London Sci. Ser. B 124:5681
109. Tibbetts EA. 2002. Visual signals of individual identity in the wasp Polistes fuscatus. Proc.
R. Soc. London Sci. Ser. B 269:142328
110. Tully T. 1996. Discovery of genes involved with learning and memory: an experimental
synthesis of Hirschian and Benzerian perspectives. Proc. Natl. Acad. Sci. USA93:1346067
111. Tully T, Cambiazo V, Kruse L. 1994. Memory through metamorphosis in normal and
mutant Drosophila. J. Neurosci. 14:6874
www.annualreviews.org Evolutionary Biology of Insect Learning 159
A
n
n
u
.

R
e
v
.

E
n
t
o
m
o
l
.

2
0
0
8
.
5
3
:
1
4
5
-
1
6
0
.

D
o
w
n
l
o
a
d
e
d

f
r
o
m

a
r
j
o
u
r
n
a
l
s
.
a
n
n
u
a
l
r
e
v
i
e
w
s
.
o
r
g
b
y

C
O
R
N
E
L
L

U
N
I
V
E
R
S
I
T
Y

o
n

0
2
/
0
8
/
0
8
.

F
o
r

p
e
r
s
o
n
a
l

u
s
e

o
n
l
y
.
ANRV330-EN53-08 ARI 2 November 2007 16:17
112. Vet LEM, Lewis WJ, Carde RT. 1995. Parasitoid foraging and learning. In Chemical
Ecology of Insects 2, ed. RT Carde, WJ Bell, pp. 65101. New York: Chapman and Hall
113. Visscher PK, Seeley TD. 1982. Foraging strategy of honeybee colonies in a temperate
deciduous forest. Ecology 63:1790801
114. von Frisch K. 1967. The Dance Language and Orientation of Bees. Cambridge, MA: Harvard
Univ. Press
115. Wackers FL, Lewis WJ. 1999. A comparison of color-, shape- and pattern-learning by
the hymenopteran parasitoid Microplitis croceipes. J. Comp. Physiol. A 184:38793
116. Waddell S, QuinnWG. 2001. Flies, genes, andlearning. Annu. Rev. Neurosci. 24:1283309
117. Wcislo WT. 1987. The role of learning in the mating biology of a sweat bee Lasioglossum
zephyrum (Hymenoptera, Halictidae). Behav. Ecol. Sociobiol. 20:17985
118. WertheimB, van Baalen E-JA, Dicke M, Vet LEM. 2005. Pheromone-mediated aggrega-
tion in nonsocial arthropods: an evolutionary ecological perspective. Annu. Rev. Entomol.
50:32146
119. West-Eberhard MJ. 2003. Developmental Plasticity and Evolution. Oxford: Oxford Univ.
Press. 794 pp.
120. Wolf R, Heisenberg M. 1991. Basic organization of operant behavior as revealed in
Drosophila ight orientation. J. Comp. Physiol. A 169:699705
121. Worden BD, Papaj DR. 2005. Flower choice copying in bumblebees. Biol. Lett. 1:5047
122. Wright WG, Kirschman D, Rozen D, Maynard B. 1996. Phylogenetic analysis of
learning-related neuromodulation in molluscan mechanosensory neurons. Evolution
50:224863
123. A report
suggesting
learning-mediated
individual
recognition in fruit
ies.
123. Yurkovic A, Wang O, Basu AC, Kravitz EA. 2006. Learning and memory associated
with aggression in Drosophila melanogaster. Proc. Natl. Acad. Sci. USA 103:17519
24
124. A study
suggesting
associative learning
in a classic simple
model animal.
124. Zhang Y, Lu H, Bargmann CI. 2005. Pathogenic bacteria induce aversive olfactory
learning in Caenorhabditis elegans. Nature 438:17984
160 Dukas
A
n
n
u
.

R
e
v
.

E
n
t
o
m
o
l
.

2
0
0
8
.
5
3
:
1
4
5
-
1
6
0
.

D
o
w
n
l
o
a
d
e
d

f
r
o
m

a
r
j
o
u
r
n
a
l
s
.
a
n
n
u
a
l
r
e
v
i
e
w
s
.
o
r
g
b
y

C
O
R
N
E
L
L

U
N
I
V
E
R
S
I
T
Y

o
n

0
2
/
0
8
/
0
8
.

F
o
r

p
e
r
s
o
n
a
l

u
s
e

o
n
l
y
.
AR330-FM ARI 9 November 2007 13:20
Annual Review of
Entomology
Volume 53, 2008 Contents
Frontispiece
Geoffrey G.E. Scudder p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p xiv
Threads and Serendipity in the Life and Research of an Entomologist
Geoffrey G.E. Scudder p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 1
When Workers Disunite: Intraspecic Parasitism by Eusocial Bees
Madeleine Beekman and Benjamin P. Oldroyd p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 19
Natural History of the Scuttle Fly, Megaselia scalaris
R.H.L. Disney p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 39
A Global Perspective on the Epidemiology of West Nile Virus
Laura D. Kramer, Linda M. Styer, and Gregory D. Ebel p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 61
Sexual Conict over Nuptial Gifts in Insects
Darryl T. Gwynne p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 83
Application of DNA-Based Methods in Forensic Entomology
Jeffrey D. Wells and Jamie R. Stevens p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 103
Microbial Control of Insect Pests in Temperate Orchard Systems:
Potential for Incorporation into IPM
Lawrence A. Lacey and David I. Shapiro-Ilan p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 121
Evolutionary Biology of Insect Learning
Reuven Dukas p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 145
Roles and Effects of Environmental Carbon Dioxide in Insect Life
Pablo G. Guerenstein and John G. Hildebrand p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 161
Serotonin Modulation of Moth Central Olfactory Neurons
Peter Kloppenburg and Alison R. Mercer p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 179
Decline and Conservation of Bumble Bees
D. Goulson, G.C. Lye, and B. Darvill p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 191
Sex Determination in the Hymenoptera
George E. Heimpel and Jetske G. de Boer p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 209
vii
A
n
n
u
.

R
e
v
.

E
n
t
o
m
o
l
.

2
0
0
8
.
5
3
:
1
4
5
-
1
6
0
.

D
o
w
n
l
o
a
d
e
d

f
r
o
m

a
r
j
o
u
r
n
a
l
s
.
a
n
n
u
a
l
r
e
v
i
e
w
s
.
o
r
g
b
y

C
O
R
N
E
L
L

U
N
I
V
E
R
S
I
T
Y

o
n

0
2
/
0
8
/
0
8
.

F
o
r

p
e
r
s
o
n
a
l

u
s
e

o
n
l
y
.
AR330-FM ARI 9 November 2007 13:20
The Argentine Ant: Challenges in Managing an Invasive
Unicolonial Pest
Jules Silverman and Robert John Brightwell p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 231
Diversity and Evolution of the Insect Ventral Nerve Cord
Jeremy E. Niven, Christopher M. Graham, and Malcolm Burrows p p p p p p p p p p p p p p p p p p 253
Dengue VirusMosquito Interactions
Scott B. Halstead p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 273
Flash Signal Evolution, Mate Choice, and Predation in Fireies
Sara M. Lewis and Christopher K. Cratsley p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 293
Prevention of Tick-Borne Diseases
Joseph Piesman and Lars Eisen p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 323
Entomological Reactions to Darwins Theory in the
Nineteenth Century
Gene Kritsky p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 345
Resource Acquisition, Allocation, and Utilization in Parasitoid
Reproductive Strategies
Mark A. Jervis, Jacintha Ellers, and Jeffrey A. Harvey p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 361
Population Ecology of Insect Invasions and Their Management
Andrew M. Liebhold and Patrick C. Tobin p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 387
Medical Aspects of Spider Bites
Richard S. Vetter and Geoffrey K. Isbister p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 409
Plant-Mediated Interactions Between Whiteies, Herbivores,
and Natural Enemies
Moshe Inbar and Dan Gerling p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 431
Ancient Rapid Radiations of Insects: Challenges for
Phylogenetic Analysis
James B. Whiteld and Karl M. Kjer p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 449
Fruit Fly (Diptera: Tephritidae) Host Status Determination: Critical
Conceptual, Methodological, and Regulatory Considerations
Martn Aluja and Robert L. Mangan p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 473
Codling Moth Management and Chemical Ecology
Peter Witzgall, Lukasz Stelinski, Larry Gut, and Don Thomson p p p p p p p p p p p p p p p p p p p p p 503
Primer Pheromones in Social Hymenoptera
Yves Le Conte and Abraham Hefetz p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 523
viii Contents
A
n
n
u
.

R
e
v
.

E
n
t
o
m
o
l
.

2
0
0
8
.
5
3
:
1
4
5
-
1
6
0
.

D
o
w
n
l
o
a
d
e
d

f
r
o
m

a
r
j
o
u
r
n
a
l
s
.
a
n
n
u
a
l
r
e
v
i
e
w
s
.
o
r
g
b
y

C
O
R
N
E
L
L

U
N
I
V
E
R
S
I
T
Y

o
n

0
2
/
0
8
/
0
8
.

F
o
r

p
e
r
s
o
n
a
l

u
s
e

o
n
l
y
.