Download When the group denies individual nutritional wisdom

ANIMAL BEHAVIOUR, 2007, 74, 931e939
doi:10.1016/j.anbehav.2006.12.022
When the group denies individual nutritional wisdom
A. DUS SUTOUR * †, S. J . S IM PSON *, E. DES PLAN D† & N. C OLAS U RDO†
*School of Biological Sciences, University of Sydney
yBiology Department, Concordia University
(Received 13 June 2006; initial acceptance 6 October 2006;
final acceptance 20 December 2006; published online 31 August 2007; MS. number: 8998R)
In social insects, decisions are usually predicted to be more accurate than individual decisions, because individual preferences are amplified at the group level. Although social caterpillars, as individuals, possess
considerable ‘nutritional wisdom’ and have the capacity to regulate their nutrient intake: does this
mean that the collective nutritional decisions made by colonies are better? We investigated individual
feeding responses and collective decisions when caterpillar colonies were faced with different food sources.
Forest tent caterpillars, Malacosoma disstria, had to choose between nutritionally balanced and unbalanced
food sources placed at opposite ends of a bridge. At the individual level, caterpillars showed a clear preference for the balanced food source. In contrast, at the collective level, colonies of caterpillars randomly
chose one of the two, and then became trapped for at least 24 h as a result of trail following behaviour.
If the first caterpillar found the unbalanced food, it was followed by the others and the colony was unable
to reverse this choice. This study provides evidence for a novel potential cost to group living, and shows
that group feeding behaviour cannot always be simply predicted from individual preferences.
Crown Copyright Ó 2007 Published by Elsevier Ltd on behalf of The Association for the Study of Animal Behaviour. All rights reserved.
Keywords: amplification; collective decision; foraging; forest tent caterpillars; Malacosoma disstria; preference
In social insects, many group tasks lead to collective
decisions, which are reliant on interactions having taken
place between individuals which have access only to local
information (Camazine et al. 2001). Collective decisions
occur when a colony is faced with several opportunities,
such as food sources with different qualities. Information
about these opportunities is gathered by individuals and
relayed to nestmates. The by-product of this distributed
information transfer is, most of the time, the collective
choice of the best alternative, without any individual necessarily being aware of all the available alternatives
(reviewed in Detrain et al. 1999).
The most common mechanisms leading to accurate
collective decisions in social insect are allelomimesis
(roughly speaking, do what my neighbour is doing; see
e.g. Sudd 1963; Altmann 1985; Deneubourg & Goss 1989)
and differential signalling. Allelomimesis, by definition,
Correspondence and present address: A. Dussutour, School of Biological
Sciences, Heydon-Laurence Building A 08, University of Sydney, NSW
2006, Australia (email: [email protected]). E. Despland and
N. Colasurdo are now at the Biology Department, Concordia University,
7141 Sherbrooke St. W, Montreal, Quebec, H4B 1R6, Canada.
leads to amplification, or a snowball effect: that is, I imitate others, others imitate me, and we all end up doing
the same thing. Amplification is an essential component
of many collective decisions observed in eusocial insects,
such as recruitment to a single food source (ants: Pasteels
et al. 1987; honeybees: Seeley et al. 1991; social caterpillars: Fitzgerald 1995). An individual that has discovered
a profitable food source conveys a signal leading to the
source. This signal is attractive to congeners and triggers
the onset of a recruitment process. Recruited individuals
follow the signal to the food source and in turn reinforce
it. After a certain amount of time the majority of the
foragers exploit the same food source. A decision emerges
because an individual behaviour is amplified by the action
of many other individuals.
Foraging in groups is assumed to be adaptive because it
can involve information sharing and better nutritional
choices (e.g. see Giraldeau & Caraco 2000; Krause & Ruxton 2002). When faced with different food sources, foragers need first to assess the nutritional quality of the
sources and then to adjust the intensity of the signal
conveyed in relation to food source quality (ants: Beckers
et al. 1990, 1992, 1993; Mailleux et al. 2000; honeybees:
Seeley et al. 1991; Seeley 1995; social caterpillars: Fitzgerald
931
0003e 3472/07/$30.00/0 Crown Copyright Ó 2007 Published by Elsevier Ltd on behalf of The Association for the Study of Animal Behaviour. All rights reserved.
932
ANIMAL BEHAVIOUR, 74, 4
1995; Ruf et al. 2001). This difference in signalling can be
amplified by the group and lead to the selection of the
nutritionally better food.
Since in social insects it has been shown that collective
decisions are usually more accurate than individual decisions because individual preferences are amplified at the
level of the group (e.g. Couzin et al. 2005; Dussutour et al.
2005), does this mean that the collective nutritional
decisions made by social species between different food
sources would automatically reflect and amplify the ‘nutritional wisdom’ of the individuals? Our aim was to address
this question using the forest tent caterpillars Malacosoma
disstria, a gregarious, outbreaking forest pest species that
uses recruitment through scent trails to exploit food sources. In referring to caterpillars as social species in this
paper, we are using the term in its broader sense as it
applies to aggregates of individuals in which there is ‘reciprocal communication of a cooperative nature’ (Wilson
1971).
Using the forest tent caterpillar allowed us to build upon
an extensive literature on nutrient regulation in other,
nonsocial caterpillars, as well as recent work on isolated
individuals of this same species (Despland & Noseworthy
2006; Noseworthy & Despland 2006). It is well established
for many caterpillar species that individuals have the capacity to regulate their intake of both protein and carbohydrate to obtain an optimal balance of nutrients from
nutritionally varied foods (e.g. Heliothis zea: Friedman et al.
1991; Heliothis virescens: Telang et al. 2001; Manduca sexta:
Thompson & Redak 2000; Spodoptera littoralis: Simpson
et al. 1988, 2004; Lee et al. 2002; Spodoptera exempta: Lee
et al. 2003).
Isolated forest tent caterpillars respond to the nutritional quality of their diet (Noseworthy & Despland 2006),
preferentially feeding from carbohydrate-enriched and
untreated trembling aspen leaves over 24-h periods and
avoiding protein-enriched leaves. Using synthetic diets,
Despland & Noseworthy (2006) found that individual forest tent caterpillars had a less marked ability to regulate intake of protein and carbohydrate than other caterpillar
species, but ate substantially more of a nutritionally
balanced diet than one deficient in either protein or carbohydrate. When in groups, caterpillars lay down a chemical
trail that stimulates and orients others in the group
(Fitzgerald & Costa 1986; Fitzgerald 1995; Ruf et al. 2001;
Colasurdo & Despland 2005). Like ants, recruited caterpillars reinforce the pheromone trails, leading to allelomimetic amplification and to the collective exploitation of
one food source. During outbreaks, defoliation by the forest tent caterpillar is not random. Specific trees and parts
of tree are attacked more often than others, which may
at least, in part, be due to variation in nutritional value
of foliage (Panzuto et al. 2001; Lévesque et al. 2002).
We carried out a series of laboratory experiments with
a simple experimental set-up, consisting of a bridge
offering the forest tent caterpillars the choice between
two food sources, which individual caterpillars are known
to be able to distinguish in nutritional value (Despland &
Noseworthy 2006): one containing a balanced ratio of
protein and carbohydrate, the other lacking digestible carbohydrate. We measured the food preference expressed by
individual caterpillars within groups, as well as the collective decisions made by the group.
METHODS
Species Studied and Rearing Conditions
The nomadic forest tent caterpillar M. disstria is a common pest of deciduous trees in North America. They are
colonial during their larval stage where colonies typically
consist of several hundred siblings that emerge from
a common egg mass. The forest tent caterpillar is the
only species of tent caterpillar that does not spin silk tents.
Instead, they spin silk mats as temporary resting sites, or
bivouacs. They also spin silk threads as they travel and
mark them with a trail pheromone by pressing the tip of
the abdomen against the substrate, much in the same
manner as ants. These pheromone trails maintain colony
cohesion. In nature, when foraging, forest tent caterpillars
travel en masse and feed simultaneously at the same site
(Fitzgerald 1995). In second instars, the distance between
the bivouac and the food source never exceeds 50 cm. After feeding to repletion, the caterpillars set off together in
search of a suitable temporary resting site to digest their
meal. Colonies display an alternation between feeding
and resting periods and use the same bivouac for several
days (Fitzgerald 1995).
Forest tent caterpillar egg masses collected in Alberta
were individually hatched and reared on a nutritionally
standard meridic artificial diet (Addy 1969) under controlled light and temperature regimes: 12:12 h photoperiod and 22 C. Experimental insects were removed from
the culture within 24 h of moulting to the second instar.
Experimental Set-up and Protocol
Artificial foods were used because they enabled us to
manipulate and standardize nutrient content. The experimental foods were based on the Addy diet, but carbohydrate content was manipulated: nutritionally balanced
food: 20% dry weight protein and 20% dry weight
carbohydrate; unbalanced food: 20% dry weight protein
and 0% dry weight carbohydrate. The protein content of
the food consisted of casein, while the carbohydrate
portion was made up of dextrose. Other components of
the food were Wesson’s salt (5.7%), cholesterol (1%),
sorbic acid (0.7%), methyl paraben (0.4%), choline chloride (0.6%) raw linseed oil (1.9%), ascorbic acid (2.9%),
sodium alginate (2.9%), antibiotic (1%) and Vanderzant
vitamin mixture (8.2%). Cellulose, a non-nutritive bulk
agent, was added to fill the remaining part of the food.
The food was presented to the insects in a 2% agar
solution.
For each replicate, an experimental group of 40 caterpillars belonging to the same egg mass was placed in the
centre of a cardboard bridge (20 3 cm; Fig. 1). The caterpillars were confined by a plastic barrier coated with Fluon
and food deprived for 2 h. After the barrier was removed
they had access to the two food sources positioned at
the ends of the bridge. Each diet was presented as a block
DUSSUTOUR ET AL.: GROUP DENIES INDIVIDUAL NUTRITIONAL WISDOM
20 cm
Balanced food
Unbalanced food
3 cm
40 Caterpillars
Figure 1. Experimental set-up. During the experiment caterpillars had access from the centre of the bridge to the two food sources differing in
quality.
of food (3 2 cm and 2 cm high) placed on a plastic
square at each end of the bridge (Fig. 1).
The bivouac area was marked with artificial pheromone
(5b-cholestan-3,24-dione) diluted in hexane to encourage
the caterpillars to come back to the centre of the bridge
after each foraging period. The whole experimental set-up
was isolated from sources of disturbance by surrounding it
with white cardboard walls. All replicates were filmed
(time lapse: one image per second) for 24 h by a video
camera placed over the bridge connected to a computer.
Twenty replicates were achieved with 20 different egg
masses.
The data were first collected at the individual level, to
quantify how individual caterpillars in a group responded
to the two food sources, and then at the collective level, to
determine the group decision.
(Simpson & Raubenheimer 2000). A meal consisted of
multiple contacts during a foraging period, which ended
when the caterpillars returned to the bivouac to rest (see
Fig. 2). Moreover, the time locomoting during the first foraging period for each food source was noted (Fig. 2), providing an indication of ‘exploratory behaviour’.
Behavioural data were collected for a total of 40 individual caterpillars observed across six replicates (egg
masses) for each food source (six or seven caterpillars out
of 40 per replicate; Table 1). Since caterpillars were not individually marked, the probability that we observed the same
caterpillar for the first and the last meal of the 24-h period
was less than 0.025, hence they were treated as statistically
independent. We conducted statistical tests (nested
ANOVA) to be sure that there was not an egg mass (replicate)
effect. The nested design allowed us to test two things: (1)
difference between unbalanced and balanced food and (2)
the variability of the egg masses within food type.
Data Collection
Individual preference
For the individual responses, we followed individuals
chosen at random from within the group and observed
them until they completed a meal. If a focal individual
was interrupted by another during the course of the meal
(which happened in fewer than 20% of cases), it was
disregarded and another caterpillar was chosen. This
procedure allowed us to be sure that feeding responses
reflected the nutritional quality of the food and not
disturbance by other caterpillars. Given the stochastic
nature of disturbance by others, and the fact that fewer
than 20% of individuals were, in fact, disturbed during the
observation, there is no reason to suspect that the dietary
responses of the focal, undisturbed larvae were not
representative of all individuals within the group.
Caterpillars were observed at the beginning of the
experiment and after 24 h. Variables recorded for focal caterpillars were the duration of the first contact with the
food (Fig. 2), which reflects taste responses to the food
(Simpson & Raubenheimer 2000), and the duration of
the first and the last meals eaten over the 24-h recording
period (Fig. 2) for each food source, which gives an approximation of the amount eaten during those meals
Collective decision
To investigate the group decision for each colony, the
number of caterpillars at the bivouac and at each food
source was measured every 10 min for 24 h for all replicates. Ten minutes was chosen because caterpillars could
not leave the bivouac, visit a food source and walk back
to the bivouac in this time. To test whether caterpillars
preferred one food source over the other (asymmetric distribution) or whether they showed no preference (symmetric distribution), we used a binomial test on the
number of caterpillars on each food source in each replicate. The null hypothesis was that caterpillars choose
both sources with equal probability (Siegel & Castellan
1988). We arbitrarily assumed that a food source was selected when the binomial test showed a significantly
higher number of foragers on one food source.
Group activity pattern
In our experiments, as in their natural environment,
forest tent caterpillars left the bivouac in search of food
and moved back to the bivouac to rest. We thus observed
an alternation between foraging and resting periods
throughout the 24 h. To investigate the effect of food
933
934
ANIMAL BEHAVIOUR, 74, 4
Second meal
First meal
(a)
Resting period
First contact
Foraging period
Time
(b)
(c)
Figure 2. (a) Foraging pattern for caterpillars. A foraging period consists of contact with food and exploring bouts. A meal is made up of the
sum of multiple contacts with food within a foraging period. Resting periods represent time spent inactive. (b) Illustration of patterns of
foraging for balanced food and (c) unbalanced food (,: resting; -: eating; : exploring).
type on the activity rhythm of the colony we measured
the number of foraging and resting periods, and their duration for all experiments for each replicate. A colony was
considered to be in a foraging state when there were more
than five caterpillars walking. A foraging period was defined to include both walking and feeding from the time
activity was initiated in the resting group until the time
the group settled again at the bivouac. It was extremely
rare that the 40 caterpillars were active at the same time.
Most of the time, 50% of the group was active (mean
number of active caterpillars during a foraging period
18.23 2.11).
All statistical tests were conducted with SPSS for Windows (version 12, SPSS Incorporated, Chicago, IL, U.S.A.).
Results are given in the text as mean SE and all the probabilities are two tailed.
P < 0.001 for length of first feeding contact, egg mass (food
type): F10,68 ¼ 0.22, P ¼ 0.991; Fig. 3a and see Fig. 2b for an
example) and for longer when they were on the nutritionally balanced food than when the food lacked carbohydrate
(food effect: F1,146 ¼ 69.943, P < 0.001, egg mass (food
type): F10,146 ¼ 0.27, P ¼ 0.987; Fig. 3c and see Fig. 2b for
an example). This difference in response to the balanced
versus unbalanced food persisted until the last meal
(Fig. 3c), although caterpillars ate more in the last than in
the first meal (time (food type): F2,146 ¼ 139.95; Fig. 3c).
The time spent locomoting (in ‘exploration’) was significantly longer for the unbalanced food than for the balanced
food (ANOVA with egg mass nested within food type, food
effect: F1,68 ¼ 45.74, P < 0.001, egg mass (food type):
F10,68 ¼ 0.36, P ¼ 0.958; Fig. 3b and see Fig. 2c for an
example).
RESULTS
Collective Decision
Individual Preference
As expected, caterpillars fed more readily (ANOVA with
egg mass nested within food type, food effect: F1,68 ¼ 65.79,
Table 1. Experimental framework used to investigate individual
preference
Egg mass
observed
Number of caterpillars
observed
Balanced food
A
B
C
E
F
G
7
7
7
7
6
6
Unbalanced
food
H
I
J
K
L
M
7
7
7
7
6
6
Total
12
80
The groups of caterpillars foraged on only one source
during the 24 h in 18 out of 20 replicates (binomial test:
P < 0.05 in these cases; Fig. 4). In 11 out of the 20 replicates caterpillars initially chose the unbalanced food
(Fig. 4). When the first caterpillar moved towards the unbalanced food it was followed by the rest of the group in
nine out of these 11 cases and a trail emerged. If the first
caterpillar moved towards the balanced source, it was
also followed by the rest of the group (seven out of nine
cases). The caterpillars were then trapped in their own trail
and, over the 24-h trial, were incapable of reversing the
choice made during the first feeding period (two reversals
were seen out of 11 initial choices for the unbalanced
food). Figure 5 illustrates one experiment where the caterpillars chose the unbalanced food source. Some caterpillars reached the balanced source during the first feeding
periods but failed to bring the group to it. Hence, the collective choice observed during the first feeding period was
random (c21 ¼ 0:89, P ¼ 0.346) and in most cases determined the pattern for the remainder of the experiment
(Fig. 4).
DUSSUTOUR ET AL.: GROUP DENIES INDIVIDUAL NUTRITIONAL WISDOM
0.5
(a)
Proportion of replicates
*
250
200
150
0.4
0.3
0.2
0.1
100
0
50
0 – 0.2
0.2 – 0.4
0.4 – 0.6
0.6 – 0.8
0.8–1
Proportion of caterpillars visiting
the balanced source
0
Figure 4. Collective decision. Frequency distribution of the proportion of caterpillars visiting the unbalanced food source (N ¼ 20) during the first feeding period (,) and during the whole experiment
(24 h; -). Proportion of caterpillars visiting the balanced source
was calculated from the sum of the 10-min periods during the first
feeding period and during the whole experiment.
First contact
(b)
*
1750
1500
Group Activity Pattern
Time (s)
1250
1000
750
500
250
0
Exploration
*
(c)
750
600
*
*
450
300
DISCUSSION
150
0
The number of foraging excursions over 24 h was significantly higher when the group concentrated its feeding on
the unbalanced than the balanced food source (t20 ¼ 3.05,
P ¼ 0.007, 9.50 0.68 and 6.62 0.53 for the unbalanced
and balanced food, respectively). Moreover, the duration
of each foraging excursion was significantly longer when
the source was nutritionally unbalanced (two-way ANOVA
with repeated measures on foraging time: food effect:
F1,18 ¼ 7.54, P ¼ 0.013, mean duration of a foraging period: 5800 351 s and 4275 430 s for the unbalanced
and balanced food, respectively). Since individual caterpillars actually took shorter meals on the unbalanced food
(see above), the difference in foraging time was due to
the substantially higher level of locomotion when foraging from the unbalanced than the balanced food source.
The first foraging period was longer than subsequent
ones (time effect: F4,72 ¼ 17.44, P < 0.001, mean duration
of the first feeding period: 11 850 1323 s and
7650 1621 s for the unbalanced and balanced food, respectively) for both of the two food sources (interaction
time*food: F4,72 ¼ 2.48, P ¼ 0.094), due to exploration
and trail formation.
First meal
Last meal
Figure 3. Individual behaviour. Duration of the first contact (a); the
exploration (b) and the first and the last meal (c) according to food
quality (95% CI). These measures were made for 40 individual caterpillars for each food type. *P < 0.001; ,: unbalanced; -: balanced
food).
As expected from earlier work (Despland & Noseworthy
2006; Noseworthy & Despland 2006), individual forest
tent caterpillars within a group ate more and spent less
time locomoting when they found themselves on the nutritionally balanced than the unbalanced food. In light of
these results and previous studies conducted with ants (see
e.g. de Biseau et al. 1991; Beckers et al. 1993), we supposed
that we would observe the same preference for the balanced food at a collective level. However, our experiments
935
ANIMAL BEHAVIOUR, 74, 4
40
35
Number of caterpillars
936
30
25
20
15
10
5
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
Time (h)
Figure 5. Number of caterpillars visiting the unbalanced food (grey lines) source and the balanced (black lines) source every 10 min during one
experiment (24 h).
showed that, in spite of a high level of preference for balanced food at an individual level, groups of caterpillars
that had initially arrived at the unbalanced food source
did not move to the balanced food, but rather became
trapped at that site for the next 24 h. This entrapment occurred despite markedly enhanced levels of locomotion
when on the unbalanced food, which, if caterpillars
were alone, would presumably have substantially increased the likelihood of encountering the balanced
food. Even when individual caterpillars did escape the
group and located the balanced food source, they were unable to recruit the rest of the group.
This result is contrary to studies conducted on eusocial
insects, where individual nutritional preferences were
considerably amplified through interactions between individuals, leading to more accurate collective decisions (ants:
see e.g. Verhaeghe 1982; de Biseau et al. 1991; Beckers
et al. 1993; Mailleux et al. 2000; honeybees: Seeley et al.
1991; Seeley 1995; termites: Andara et al. 2004). To
achieve a decision when faced with several alternatives,
bees and ants assess the profitability of food sources and
adjust the intensity of the signal conveyed. Bees modulate
the strength of the waggle dance and ants modulate the
quantity of pheromones deposited. The collective decisions observed in our experiments showed that caterpillars are not able to transmit the information about food
quality to their congeners even if they are able to assess
it. During exploration, the first individual left the bivouac
area and randomly chose a direction while laying down
a trail. It was then followed by the other caterpillars,
which in turn, reinforced the trail. If the first caterpillar
found the unbalanced food, it was joined by the rest of
the group. After feeding, the group came back to the bivouac, reinforcing the trail. During subsequent feeding periods the caterpillars followed the trail and the group was
not able to reverse its choice even if individuals were able
to find the balanced source. Caterpillars seem to lack flexibility in trail laying and trail following behaviour, which
usually allows the formation of new trails (Deneubourg
et al. 1983; Fitzgerald & Costa 1986). The difference observed between individual preference and group choice
in caterpillars resulted from an excessively strong group
cohesive mechanism. In a famous essay, Fabre (1899) reported that processionary caterpillars could be trapped
in their own trail for more than 8 days. In social caterpillars, trail laying and trail following promote individuals remaining together, which contribute significantly to
individual fitness (Robison 1993). The price of such high
group cohesion is suboptimal diet choices, because individuals are unable to use this system to transmit information about food quality.
Lack of flexibility during foraging was also shown in
ants when the food sources were not offered at the same
time. When the richer source is discovered after the poorer
one, the richer source is only weakly exploited (Pasteels
et al. 1987; Beckers et al. 1990; Traniello & Robson
1995). However, in ants experiments were conducted
with less extremely biased diet, that is, the poor food
source still contained carbohydrate, which is an essential
nutrient. Also, these experiments were conducted over
short periods of time (2 h), and after a few more hours
ants were able to reverse their choice (A. Dussutour, personal observations). Collective behaviour needs a balance
between efficiency in making group decisions and flexibility which allows reversal of decisions. Contrary to eusocial
insects, caterpillars lack such flexibility.
It was evident, therefore, that forest tent caterpillars
could become trapped on a nutritionally unbalanced food
source, but two key questions remain. First, does becoming trapped constitute a fitness cost; and second, are
groups of caterpillars more readily trapped than individual
caterpillars? Regarding the first of these issues, there is
little doubt that a diet entirely deficient in carbohydrate is
deleterious to fitness. In caterpillars, as for other phytophagous insects (Raubenheimer & Simpson 1999), the role of
carbohydrates as essential nutrients has been emphasized
in numerous studies. Survival rate, survival time and development rate are low for caterpillars reared with diets
DUSSUTOUR ET AL.: GROUP DENIES INDIVIDUAL NUTRITIONAL WISDOM
lacking carbohydrates (Harvey 1974; Clancy 1992; Lee
et al. 2002, 2003; M. disstria: Colasurdo 2006; Despland
& Noseworthy 2006), with implications for adult fecundity (Honek 1993; M. disstria: Colasurdo 2006) and firstinstar survival of the progeny (Carisey & Bauce 2002). A
majority of the caterpillar species listed by Simpson &
Raubenheimer (1993) required 50e60% protein and
50e40% carbohydrate in artificial diets for optimal
development. Carbohydrates, as essential nutrients, are
perceived as phagostimulants by most insects studied
(Panzuto et al. 2001, 2002).
Regarding the second question, there is little doubt that
being in a group impeded individual caterpillars from
locating the balanced food once the initial choice had
been made for the unbalanced food source. Increased
exploration in response to nutrient deprivation is commonly observed as a functional strategy to find something
better (Morris & Kareiva 1990; Barton-Brown 1993). Caterpillars in our study showed increased locomotory (‘exploratory’) activity when on the suboptimal food. As a result,
a number of individuals did, in fact, locate the balanced
food (e.g. Fig. 5), as compared to none on the unbalanced
food in cases where the group arrived initially at the balance food source. But in our experiment even if an individual was able to find the balanced source, it was
unable to bring the other caterpillars because they remained trapped in the first trail formed. Thus, even
when the group spent more time locomoting, it did not
find the balanced food source within 24 h. It should also
be mentioned that, in nature, feeding more frequently
and increasing exploratory behaviour increases predation
risk, providing another cost to being restricted to an unbalanced food in addition to those discussed above (Krebs
& Davies 1993; Bernays 1997). Consequently, the selective
foraging on certain parts of the tree observed by Lévesque
et al (2002) could be the result of a difference in survival
rate and not the expression of a group preference.
The feeding behaviour of colonies that were trapped on
the unbalanced food suggested a vain attempt to
compensate for their incorrect choice by feeding more
frequently on the nutritionally unbalanced food. Compensation in response to being restricted to nutritionally
poor food has been extensively shown in insects (e.g.
Simpson & Simpson 1990; Simpson & Raubenheimer
2000; Lee et al. 2004).
CONCLUSION
Our results emphasize the difference between individual
behaviour and group decisions. They underline the fact
that it is the information transmitted by individuals
combined with the amplification phenomena that govern
the group choice and not the individual preference.
Previous work shows that using group decisions to infer
individual preference often overestimates individual selectivity (Dussutour et al. 2005). We show here that the
opposite can also be true: individuals can be more selective than groups. Together these findings emphasize that
a direct relationship between individual preference and
group choice cannot always be assumed. These findings
could be relevant for the study of other group-living
species, including not only social insects like ants, wasps
and termites, but also cockroaches, other gregarious caterpillars, locusts, sawflies, and bark beetles. Animals’ feeding
preferences are often studied as a means of predicting population distribution or feeding damage. In many experiments individuals are tested alone and the effect of
amplifying mechanisms through the exchange of chemical, visual, tactic or acoustic signals between individuals
are rarely considered (caterpillars: Simmonds et al. 1992;
Stamp & Casey 1993; Alborn et al. 1996; Showler 2001;
Lazarevic et al. 2003; Ghumare & Mukherjee 2005; Miles
et al. 2005; cockroaches: Carrel & Tanner 2002; Jones &
Raubenheimer 2002; Peterson et al. 2002; sawflies:
Mansfield & Mills 2002; beetles: Poland et al. 2003). As a
consequence, the level of selectivity of group feeding can
be underestimated (if communication amplifies individual
preferences) or overestimated (if, as in our study, communication traps groups following random initial discoveries). It is consequently important to understand how the
group chooses between different resources, and not solely
base our knowledge on individual preference.
Acknowledgments
We thank the Behavioural ecology group of the Biology
Department at Concordia University for their comments
on the manuscript. A.D. was supported by postdoctoral
grants from foundation Fyssen and from Singer-Polignac.
We thank Natural Sciences and Engineering Research
Council and Canadian Fund for Innovation for funding
and Dr Barry Cooke of the Canadian Forest Service for
collecting the egg masses.
References
Addy, N. D. 1969. Rearing the forest tent caterpillar on an artificial
diet. Journal of Economic Entomology, 62, 270e271.
Alborn, H. T., Röse, U. S. R. & McAuslane, H. J. 1996. Systemic
induction of feeding deterrents in cotton plants by feeding
of Spodoptera spp. larvae. Journal of Chemical Ecology, 22,
919e932.
Altmann, D. 1985. The effects of allelomimetic behavior in animals.
Proceedings of the 19th International Ethology Conference, Toulouse.
pp. 112e113.
Andara, C., Issa, S. & Jaffé, K. 2004. Decision making systems in recruitment to food for two Nasulitermitinae (Isoptera: Termitidae).
Sociobiology, 44, 1e13.
Barton-Brown, L. 1993. Physiologically induced changes in resourceoriented behavior. Annual Review of Entomology, 38, 1e25.
Beckers, R., Deneubourg, J.-L., Goss, S. & Pasteels, J.-M. 1990.
Collective decision making through food recruitment. Insectes
Sociaux, 37, 258e267.
Beckers, R., Deneubourg, J.-L. & Goss, S. 1992. Trail laying
behaviour during food recruitment in the ant Lasius niger (L.). Insectes Sociaux, 39, 59e72.
Beckers, R., Deneubourg, J.-L. & Goss, S. 1993. Modulation of trail
laying in the ant Lasius niger (Hymenoptera: Formicidae) and its
role in the collective selection of a food source. Journal of Insect
Behavior, 6, 751e759.
Bernays, E. A. 1997. Feeding by lepidopteran larvae is dangerous.
Ecological Entomology, 22, 121e123.
937
938
ANIMAL BEHAVIOUR, 74, 4
de Biseau, J.-C., Deneubourg, J.-L. & Pasteels, J.-M. 1991. Collective flexibility during food recruitment in the ant Myrmica sabuleti
(Hymenoptera: Formicidae). Psyche, 98, 323e336.
Camazine, S., Deneubourg, J.-L., Franks, N. R., Sneyd, J.,
Theraulaz, G. & Bonabeau, E. 2001. Self-organization in Biological
Systems. Princeton, New Jersey: Princeton University Press.
Carisey, N. & Bauce, E. 2002. Does nutrition-related stress carry over to
spruce budworm, Choristoneura fumiferana (Lepidoptera: Tortricidae)
progeny? Bulletin of Entomological Research, 92, 101e108.
Carrel, J. E. & Tanner, E. M. 2002. Sex-specific food preferences in
the Madagascar hissing cockroach Gromphadorhina portentosa
(Dictyoptera: Blaberidae). Journal of Insect Behavior, 15, 707e714.
Clancy, K. M. 1992. The role of sugars in western spruce budworm
nutritional ecology. Ecological Entomology, 17, 189e197.
Colasurdo, N. 2006. Nutritional performance and feeding behavior
of the forest tent caterpillar (Malacosoma disstria), Masters thesis,
Concordia University, Canada.
Colasurdo, N. & Despland, E. 2005. Social cues and following behavior
in the forest tent caterpillar. Journal of Insect Behavior, 18, 77e87.
Couzin, I. D., Krause, J., Franks, N. R. & Levin, S. A. 2005. Effective
leadership and decision making in animal groups on the move.
Nature, 433, 513e516.
Deneubourg, J.-L. & Goss, S. 1989. Collective patterns and
decision-making. Ethology, Ecology and Evolution, 1, 295e311.
Deneubourg, J.-L., Pasteels, J.-M. & Verhaeghe, J.-C. 1983.
Probabilistic behaviour in ants: a strategy of errors? Journal of
Theoretical Biology, 105, 259e271.
Despland, E. & Noseworthy, M. 2006. How well do specialist
feeders regulate nutrient intake? Evidence from a gregarious
tree-feeding caterpillar. Journal of Experimental Biology, 209,
1301e1309.
Detrain, C., Deneubourg, J.-L. & Pasteels, J.-M. 1999. Information
Processing in Social Insects. Basel: Birkhäuser-Verlag.
Dussutour, A., Deneubourg, J.-L. & Fourcassié, V. 2005. Amplification of individual preferences in a social context: the case of wallfollowing in ants. Proceedings of the Royal Society of London, Series
B, 272, 705e714.
Fabre, J.-H. 1989. Souvenirs Entomologiques. Série IV. Paris:
Delagrave.
Fitzgerald, T. D. 1995. The Tent Caterpillars. Ithaca, New York:
Cornell University Press.
Fitzgerald, T. D. & Costa, J. T. 1986. Trail-based communication
and foraging behavior of young colonies of the forest tent caterpillar Malacosoma disstria Hubn. (Lepidoptera: Lasiocampidae). Annals of the Entomological Society of America, 79, 999e1007.
Friedman, S., Waldbauer, G. P., Eertmoed, J. E., Naeem, M. &
Ghent, A. W. 1991. Blood trehalose levels have a role in the
control of dietary self-selection by Heliothis zea larvae. Journal of
Insect Physiology, 37, 919e928.
Ghumare, S. S. & Mukherjee, S. N. 2005. Absence of food aversion
learning in the polyphagous noctuid, Spodoptera litura (F.) following intoxication by deleterious chemicals. Journal of Insect Behavior, 18, 105e114.
Giraldeau, L. A. & Caraco, T. 2000. Social Foraging Theory. New
Jersey: Princeton University Press.
Harvey, G. T. 1974. Nutritional studies of eastern spruce budworm
(Lepidoptera: Tortricidae). I. Soluble sugars. Canadian Entomologist, 106, 353e365.
Honek, A. 1993. Intraspecific variation in body size and fecundity in
insects: a general relationship. Oikos, 66, 483e492.
Jones, S. A. & Raubenheimer, D. 2002. Short-term responses by the
German cockroach, Blattella germanica, to insecticidal baits: behavioural observations. Entomologia Experimentalis et Applicata,
102, 1e11.
Krause, J. & Ruxton, G. D. 2002. Living in Groups. New York: Oxford
University Press.
Krebs, J. R. & Davies, N. B. 1993. An Introduction to Behavioural Ecology. Oxford: Blackwell Scientific.
Lazarevic, J., Peric-Mataruga, V., Prolic, Z. & Tucic, N. 2003. Behavioural response to an unsuitable host plant in the gypsy
moth (Lymantria dispar L.). Folia Biologica-Krakov, 51, 129e131.
Lee, K. P., Behmer, S. T., Simpson, S. J. & Raubenheimer, D.
2002. A geometric analysis of nutrient regulation in the generalist
caterpillar Spodoptera littoralis (Boisduval). Journal of Insect Physiology, 48, 655e665.
Lee, K. P., Raubenheimer, D., Behmer, S. T. & Simpson, S. J. 2003.
A correlation between macronutrient balancing and insect hoste
plant range: evidence from the specialist caterpillar Spodoptera
exempta (Walker). Journal of Insect Physiology, 49, 1161e1171.
Lee, K. P., Raubenheimer, D. & Simpson, S. J. 2004. The effect of
nutritional imbalance on compensatory feeding for cellulosemediated dietary dilution in a generalist caterpillar. Physiological
Entomology, 29, 108e117.
Lévesque, K. R., Fortin, M. & Mauffette, Y. 2002. Temperature and
food quality effects on growth, consumption and post-ingestive
utilization efficiencies of the forest tent caterpillar Malacosoma
disstria (Lepidoptera: Lasiocampidae). Bulletin of Entomological Research, 92, 127e136.
Mailleux, A. C., Deneubourg, J.-L. & Detrain, C. 2000. How do
ants assess food volume? Animal Behaviour, 59, 1061e1069.
Mansfield, S. & Mills, N. J. 2002. Direct estimation of the survival
time of commercially-produced adult Trichogramma platneri under
field conditions. Biological Control, 29, 332e340.
Miles, C. I., del Campo, M. L. & Renwick, J. A. A. 2005. Behavioral
and chemosensory responses to a host recognition cue by larvae of
Pieris rapae. Journal of Comparative Physiology A, 191, 147e155.
Morris, W. & Kareiva, P. 1990. How insect herbivores find suitable
host plants: the interplay between random and non-random movement in insecteplant interactions. In: InsectePlant Interactions. vol.
3 (Ed. by E. Bernays), pp. 175e208. Boca Raton, Florida: CRC Press.
Noseworthy, M. & Despland, E. 2006. How do primary nutrients
affect the performance and preference of forest tent caterpillars
on trembling aspen? Canadian Entomologist, 138, 367e375.
Panzuto, M., Lorenzetti, F., Mauffette, Y. & Albert, P. J. 2001. Perception of aspen, and sun/shade sugar maple leaf soluble extracts
by larvae of Malacosoma disstria. Journal of Chemical Ecology, 27,
1e16.
Panzuto, M., Mauffette, Y. & Albert, P. J. 2002. Developmental,
gustatory, and behavioral responses of leafroller larvae, Choristoneura rosaceana, to tannic acid and glucose. Journal of Chemical
Ecology, 28, 145e160.
Pasteels, J.-M., Deneubourg, J.-L. & Goss, S. 1987. Selforganization mechanisms in ant societies (I): trail recruitment to
newly discovered food sources. In: Experientia Supplementum.
vol. 54, From Individual to Collective Behaviour in Social Insects
(Ed. by J.-M. Pasteels & J.-L. Deneubourg), pp. 155e175. Basel:
Birkhäuser-Verlag.
Peterson, C. J., Nemetz, L. T., Jones, L. M. & Coat, J. R. 2002. Behavioral activity of catnip (Lamiaceae) essential oil components to
the German cockroach (Blattodea: Blattellidae). Journal of Economic Entomology, 95, 377e380.
Poland, T. M., de Groot, P., Burke, S., Wakerchuk, D., Haack, R.,
Nott, R. & Scarr, T. 2003. Development of an improved attractive
lure for the pine shoot beetle, Tomicus piniperda (Coleoptera: Scolytidae). Agricultural and Forest Entomology, 5, 293e300.
Raubenheimer, D. & Simpson, S. J. 1999. Integrating nutrition:
a geometric approach. Entomologia Experimentalis et Applicata,
91, 67e82.
DUSSUTOUR ET AL.: GROUP DENIES INDIVIDUAL NUTRITIONAL WISDOM
Robison, D. J. 1993. The feeding ecology of the forest tent caterpillar, Malacosoma disstria Hübner, among hybrid poplar clones, Populus spp. Ph.D. thesis, University of Wisconsin.
Ruf, C., Costa, J. T. & Fiedler, K. 2001. Trail-based communication
in social caterpillars of Eriogaster lanestris (Lepidoptera: Lasiocampidae). Journal of Insect Behavior, 14, 229e245.
Seeley, T. D. 1995. The Wisdom of the Hive: the Social Physiology of
Honey Bee. Cambridge, Massachusetts: Harvard University Press.
Seeley, T. D., Camazine, S. & Sneyd, J. 1991. Collective decisionmaking in honey bees: how colonies choose among nectar sources. Behavioral Ecology and Sociobiology, 28, 277e290.
Showler, A. T. 2001. Spodoptera exigua oviposition and larval feeding preferences for pigweed, Amaranthus hybridus, over squaring
cotton, Gossypium hirsutum, and a comparison of free amino acids
in each host plant. Journal of Chemical Ecology, 27, 2013e2028.
Siegel, S. & Castellan, N. J. J. 1988. Nonparametric Statistics for the
Behavioral Sciences. New York: McGraw-Hill.
Simmonds, M. S. J., Simpson, S. J. & Blaney, W. M. 1992. Dietary
selection behaviour in Spodoptera littoralis: the effects of conditioning diet and conditioning period on neural responsiveness and selection behaviour. Journal of Experimental Biology, 162, 73e90.
Simpson, S. J. & Raubenheimer, D. 1993. A multiple level analysis of
feeding behaviour: the geometry of nutritional decisions. Philosophical Transactions of the Royal Society of London, Series B, 342, 381e402.
Simpson, S. J. & Raubenheimer, D. 2000. The hungry locust. Advances in the Study of Behavior, 29, 1e44.
Simpson, S. J. & Simpson, C. L. 1990. The mechanisms of nutritional compensation by phytophagous insects. In: InsectePlant
Interactions. vol. 2 (Ed. by E. A. Bernays), pp. 111e160. Boca Raton,
Florida: CRC Press.
Simpson, S. J., Simmonds, M. S. J. & Blaney, W. M. 1988. A
comparison of dietary selection behaviour in larval Locusta
migratoria and Spodoptera littoralis. Physiological Entomology, 13,
228e238.
Simpson, S. J., Sibly, R. M., Lee, K. P., Behmer, S. T. &
Raubenheimer, D. 2004. Optimal foraging when regulating
intake of multiple nutrients. Animal Behaviour, 68, 1299e1311.
Stamp, N. E. & Casey, T. M. 1993. Caterpillars: Ecological and Evolutionary Constraints on Foraging. New York: Chapman & Hall.
Sudd, J. H. 1963. How insects work in groups. Discovery, London, 24,
15e19.
Telang, A., Booton, V., Chapman, R. F. & Wheeler, D. E. 2001.
How female caterpillars accumulate their nutrient reserves. Journal
of Insect Physiology, 47, 1055e1064.
Thompson, S. N. & Redak, R. A. 2000. Interactions of dietary protein and carbohydrate determine blood sugar level and regulate
nutrient selection in the insect Manduca sexta L. Biochimica et Biophysica Acta, 1523, 91e102.
Traniello, J. F. A. & Robson, S. K. 1995. Trail and territorial communication in social insects. In: The Chemical Ecology of Insect. vol. 2
(Ed. by W. J. Bell & R. Cardé), pp. 241e246. London: Chapman
& Hall.
Verhaeghe, J. C. 1982. Food recruitment in Tetramorium caespitum
(Hymenoptera: Formicidae). Insectes Sociaux, 29, 67e85.
Wilson, E. O. 1971. The Insect Societies. Cambridge: Harvard
University Press.
939