ArticlePDF AvailableLiterature Review

Sick ants become unsociable


Sick ants become unsociable

Abstract and Figures

Parasites represent a severe threat to social insects, which form high-density colonies of related individuals, and selection should favour host traits that reduce infection risk. Here, using a carpenter ant (Camponotus aethiops) and a generalist insect pathogenic fungus (Metarhizium brunneum), we show that infected ants radically change their behaviour over time to reduce the risk of colony infection. Infected individuals (i) performed less social interactions than their uninfected counterparts, (ii) did not interact with brood anymore and (iii) spent most of their time outside the nest from day 3 post-infection until death. Furthermore, infected ants displayed an increased aggressiveness towards non-nestmates. Finally, infected ants did not alter their cuticular chemical profile, suggesting that infected individuals do not signal their physiological status to nestmates. Our results provide evidence for the evolution of unsociability following pathogen infection in a social animal and suggest an important role of inclusive fitness in driving such evolution.
Content may be subject to copyright.
Sick ants become unsociable
*Centre for Social Evolution, Department of Biology, University of Copenhagen, Copenhagen, Denmark
Biology Department, Emory University, Atlanta, GA, USA
àCentre for Social Evolution, Department of Agriculture and Ecology, University of Copenhagen, Copenhagen, Denmark
§Laboratoire d’Ethologie Expe
´rimentale et Compare
´e (LEEC), Universite
´Paris 13, Villetaneuse, France
Living in groups confers several advantages to the
individuals involved, including increased foraging effi-
ciency, cooperative brood care, enhanced reproduction
and predator defence (Maynard Smith & Szathmary,
1995; Bourke, 2011). Group living, however, is also
subject to significant costs mostly in the form of increased
disease risk because of higher frequencies of social
interactions and typically a high level of relatedness
within the group (Bull et al., 1991; Myers & Rothman,
1995; Schmid-Hempel, 1998; McCallum et al., 2001;
Boomsma et al., 2005). In response to parasite (broadly
defined here to include both macroparasites and micro-
parasites such as fungi) pressure, group living organisms
have evolved several lines of defence. In addition to their
individual physiological immune mechanisms, they are
able to engage in a group-level immunity consisting of a
series of behaviours that reduce the risk of infection or
minimize the parasite-induced fitness losses (Cremer
et al., 2007; Cremer & Sixt, 2009; Wilson-Rich et al.,
2009). In social insects, group-level defences include
avoidance behaviours (Oi & Pereira, 1993; Cremer et al.,
2007; Cotter & Kilner, 2010), hygienic behaviours
(reviewed in Wilson-Rich et al., 2009), collection of
antimicrobial resin (Christe et al., 2003; Castella et al.,
2008; Simone et al., 2009), spread of metapleural gland
secretions (Yek & Mueller, 2010), behavioural fever
(Starks et al., 2000) and allo-grooming (Yanagawa &
Shimizu, 2007; Walker & Hughes, 2009).
Allo-grooming is a behaviour found in many social
animals, from primates to social insects, and is a mean of
removing parasites from another individual. This social
interaction is involved in disease resistance in at least two
ways. First, allo-grooming seems to be a very efficient
way of physically removing infective fungal conidia from
nestmates (Rosengaus et al., 1998; Hughes et al., 2002;
Yanagawa et al., 2008; Walker & Hughes, 2009). Second,
allo-grooming and other social contacts have been shown
to boost host resistance to parasites by ‘social transfer’ of
disease resistance (Traniello et al., 2002; Ugelvig &
Cremer, 2007).
Recently, it has been shown that moribund ants
(Heinze & Walter, 2010) and honeybees (Rueppell et al.,
2010) leave their nest and die in isolation. Previously, it
has been found that infected ants spend less time with
the brood (Ugelvig & Cremer, 2007). These results could
Correspondence: Nick Bos, Centre for Social Evolution, Department of
Biology, University of Copenhagen, Copenhagen, Denmark.
Tel.: +45 35 32 12 38; fax: +45 35 32 12 50;
ª2011 THE AUTHORS. J. EVOL. BIOL. 25 (2012) 342–351
chemical communication;
host-parasite interaction;
life history evolution.
Parasites represent a severe threat to social insects, which form high-density
colonies of related individuals, and selection should favour host traits that
reduce infection risk. Here, using a carpenter ant (Camponotus aethiops) and a
generalist insect pathogenic fungus (Metarhizium brunneum), we show that
infected ants radically change their behaviour over time to reduce the risk of
colony infection. Infected individuals (i) performed less social interactions
than their uninfected counterparts, (ii) did not interact with brood anymore
and (iii) spent most of their time outside the nest from day 3 post-infection
until death. Furthermore, infected ants displayed an increased aggressiveness
towards non-nestmates. Finally, infected ants did not alter their cuticular
chemical profile, suggesting that infected individuals do not signal their
physiological status to nestmates. Our results provide evidence for the
evolution of unsociability following pathogen infection in a social animal
and suggest an important role of inclusive fitness in driving such evolution.
doi: 10.1111/j.1420-9101.2011.02425.x
be predicted by inclusive fitness theory, where a mori-
bund ant increases its inclusive fitness by leaving the
nest, to prevent nestmates and brood from getting
infected. Although the behavioural changes observed in
social insects following parasite exposure and infection
are likely to provide fitness benefits to the colony, and
hence represents host adaptations, they could also be the
consequence of parasitic manipulation whereby parasites
modify the host phenotype (e.g. behaviour) in a way that
increase their transmission (Schmid-Hempel, 1998;
Moore, 2002; Andersen et al., 2009; Lefevre et al.,
2009). For example, the fungal parasite Ophiocordyceps
manipulates the behaviour of its ant host causing it to
abandon its colony, and bite and cling onto the underside
of leaves where temperature and humidity conditions are
optimal for fungal growth (Andersen et al., 2009). An
increasing number of studies on generalist pathogens
provide evidence in favour of behavioural modifications
as host adaptations (Rosengaus et al., 1998; Hughes et al.,
2002; Traniello et al., 2002; Ugelvig & Cremer, 2007). For
instance, a recent study by Heinze & Walter (2010)
elegantly ruled out the alternative hypothesis of parasitic
manipulation by exposing healthy ants workers to CO
hence reducing their prospective lifespan and observing
that they desert their nest and die in seclusion exactly
like their nestmates infected by a fungus.
The behavioural changes performed by uninfected
colony members towards their infected conspecifics (e.g.
exclusion, intensive care, allo-grooming) could be based
on possible changes in the recognition cues of infected
individuals, although this has not yet been tested.
Communication in social insects occurs predominantly
through the use of chemicals. In ants, cuticular hydro-
carbons play a major role in a wide range of communi-
cation and interaction processes including nestmate and
caste recognition (reviewed in van Zweden & d’Ettorre,
2010) and home range markings (Lenoir et al., 2009).
Thus, we might expect that the cuticular hydrocarbon
profile of infected individuals would undergo a modifi-
cation triggering a change in the behaviour of their
nestmates, although other chemicals might be involved.
However, behavioural changes observed in infected
individuals, such as self-exclusion from the colony, do
not necessarily have to involve chemical communication.
To date, the different behavioural mechanisms used by
social insects to fight their parasites have mostly been
studied in isolation (but see Wilson-Rich et al., 2007). For
instance, Walker & Hughes (2009) focused on allo-
grooming, whereas Heinze & Walter (2010) focused on
ant location. Here, we present an integrated study,
investigating social interactions and chemical recognition
cues in the ant Camponotus aethiops following infection.
We exposed experimental colonies of C. aethiops to
Metarhizium brunneum, a generalist entomopathogenic
fungus commonly used as an experimental parasite of
ants and termites (Hughes & Boomsma, 2004; Calleri
et al., 2006; Chapuisat et al., 2007). First, we recorded the
location and behaviours exhibited by infected and unin-
fected ants until the death of infected individuals.
Second, we investigated whether the cuticular hydrocar-
bon profile of infected individuals were different from
that of uninfected nestmates. Finally, we tested whether
infection changes the behaviour of ants with regard to
nestmate recognition. We hypothesized that terminally
sick individuals might become more aggressive against
non-nestmates, as they have ‘nothing to lose’, and by
repelling potential colony invaders inclusive fitness
benefits could still be gained.
Study organisms
Five queenright colonies of C. aethiops (Latr.) were
collected in April 2009 in the Italian Apennines. Each
colony was housed in a Fluonª(De Monchy, the
Netherlands)-coated plastic box (27 ·17 ·9.5 cm) with
a plaster floor serving as the nest area. This was
connected with a short plastic tube to another plastic
box (27 ·17 ·9.5 cm) serving as a foraging arena. The
ants were fed twice a week with diluted honey and
mealworms (Tenebrio molitor); water was provided ad
libitum. The nests were kept in a climate room, at
25 ± 2 C, D:L 12:12 h. Subcolonies were created
2 weeks before the start of the experiment, by housing
50 ants (40 minor and 10 major workers, taken all from
the foraging arena and thus presumably of similar age)
and 10 larvae in a plastic box (17 ·11 ·6 cm) with a
plaster floor, serving as a nest. The lid of the nest box was
covered with opaque tape to darken the nest area. The
nest box was connected to another plastic box of the
same size, serving as a foraging arena, also closed with a
lid perforated with 10 holes of 2 mm diameter each.
Relative humidity inside the boxes was ± 52%. For
experiment 1, two subcolonies (test and control) were
created from each of the five colonies, for a total of 10
subcolonies. For experiments 2 and 3, four subcolonies
(two tests and two controls) were created from each of
three different colonies, for a total of 12 subcolonies.
Conidia suspensions of M. brunneum (strain KVL 04-57,
formerly named Metarhizium anisopliae, Bischoff et al.,
2009) were created according to Hughes et al. (2002).
Conidia were harvested from SDA agar plates and diluted
in 0.05% Triton-X. Their viability was checked by plating
100 lLof1·10
conidia per millilitre solution on SDA
agar plates, incubating for 24 h at 23 C and recording
germination at a·400 magnification (the per cent of
conidia germination was 98%). The conidia solution was
diluted to a final concentration of 1 ·10
viable conidia
per millilitre, assuring effective infection (c.f. Rosengaus
et al., 1998). Ants to be treated (minor workers) were
Sick ants become unsociable 343
ª2011 THE AUTHORS. J. EVOL. BIOL. 25 (2012) 342–351
marked individually with dots of enamel paint. The day
after, individual marked ants were exposed to M. brun-
neum (hereafter ‘treated ants’), by submerging them for
5 s in the conidia suspension (previously vortexed). After
this, each ant was allowed to walk on a clean filter paper
to remove excess liquid. The ant was then put back in the
foraging arena of its subcolony. For the control, individ-
ual marked ants were submerged into 0.05% Triton-X
solution (hereafter ‘control ants’).
In summary, for experiment 1, five ants per subcolony
were marked and treated. Half of the subcolonies (n=5)
were treated with conidia (test) and the other half
(n= 5) with Triton-X (control), for a total of 25 treated
and 25 control ants. For experiments 2 and 3, eight ants
were marked and treated per subcolony. Six subcolonies
were treated with conidia (tests) and the other six with
Triton-X (control). Three of these ants per subcolony
were used for experiment 2, whereas the other five were
used for experiment 3, for a total of 48 treated and 48
control ants.
Experiment 1: Behavioural study
During 6 days, starting on the day after exposure to
M. brunneum, the behaviour of each marked ant (5 ants for
each of the 10 subcolonies) was recorded for 4 min twice a
day, once in the morning (between 9 and 12
) and once
in the afternoon (between 2 and 5
), and the order in
which the different subcolonies were observed was ran-
domized. The following behaviours performed by other
workers towards the marked ants (either treated or
control) were recorded using the software E
(Ottoni, 2000): (i) No contact, (ii) Antennation, (iii) Allo-
grooming, (iv) Stomodeal trophallaxis (exchange of liquid
food). No aggressive acts (mandible opening, biting and
gaster flexing) were observed. Additionally, the following
behaviours performed by the marked ant itself were
recorded: (i) self-grooming, (ii) allo-grooming, and (iii)
stomodeal trophallaxis. As it was not always clear in which
direction trophallaxis took place, both ‘donating’ and
‘receiving’ trophallaxis were pooled together for analysis.
The location of the ants was recorded, divided into
three categories: (i) nest (excluding brood pile), (ii)
foraging arena, and (iii) brood pile. All observations were
done blind with respect to treatment. On the sixth day of
observation, only two of the infected ants were still alive,
and thus only the results from the first 5 days of
observation are reported.
When marked cadavers of both treated and control ants
were found, they were surface sterilized for 5 s in 10%
sodium hypochlorite (Lacey & Brooks, 1997) and placed in
individual medicine cups containing a ball of moist cotton
wool. These cadavers were checked for the appearance of
M. brunneum conidia to make sure the cause of death was
indeed infection. Six of 25 treated ants escaped from the
subcolonies or had their colour mark rubbed off. Of the 19
remaining focal ants that had been treated, a total of 15
(79%) became infected and none of the control ants were
found infected. All subsequent analyses are restricted to
these infected (n= 15) and control (n= 25) ants. Overall,
analyses of ant behaviours and locations were based on
248 four-minute observations of control individuals and
127 four-minute observations of treated individuals
(because infected individuals died over the course of the
Experiment 2: Chemical analysis
It takes 2–4 days for the fungus to establish an infection
and involves attachment of the conidia to the cuticle,
conidia germination, formation of appressoria and pene-
tration through the insect cuticle by enzymatic degrada-
tion and mechanical pressure (Gillespie et al., 2000). Thus,
on day three after inoculation, three marked ants per
subcolony (n= 6 test and 6 control subcolonies) were put
in individual vials and frozen. Chemical extracts were
made by submerging each ant in 150 lL of solvent
(pentane; Sigma-Aldrich, Copenhagen, Denmark) for
10 min, after having removed the gaster, which contained
the marking. The extract was transferred to a 200-lL glass
insert, the solvent was allowed to evaporate and the
extract was re-dissolved in 50 lL of solvent. Gas chroma-
tography–mass spectrometry (GC-MS) was performed as
described in Van Zweden et al. (2009), using a Agilent
Technologies (Horsholm, Denmark) 6890N gas chromato-
graph (capillary column: Agilent HP-5MS, 30 m ·
25 lm·0.25 lm; split–splitless injector; carrying gas:
helium at 1 mL min
) coupled with an Agilent 5975
mass selective detector and 70eV electron impact ioniza-
tion. We obtained chemical profiles very similar to those
found in the study of Van Zweden et al. (2009), and
identified and integrated 36 GC peaks and used them for
statistical analysis. Integration of chemical data was
performed blind with respect to treatment. Before day
three, 2 of 36 treated ants died, making a total of 34 ants
Before the start of the experiments, a pilot study was
conducted to establish whether treatment itself (sub-
merging the ant in test and control solution) affects the
chemical profile of ants. Five ants were treated with the
conidia solution, five with Triton-X, and five were not
treated at all. After 1 hour, chemical extracts were
prepared and injected in a GC-MS (same method as
above) for chemical analysis. Treatment itself (conidia
solution or Triton-X) did not affect the chemical profile
compared with nontreated ants (
, Pillai V = 1.22,
= 1.36, P= 0.30).
Experiment 3: Aggression tests
On day three after treatment, aggression tests were
conducted by placing a treated or control ant (focal
individual) and a non-nestmate (target individual) into a
circular neutral arena (50 ·60 mm), where they were
344 N. BOS ET AL.
ª2011 THE AUTHORS. J. EVOL. BIOL. 25 (2012) 342–351
allowed to habituate for 1 min. The target ant was
separated from the focal ant by a plastic cylinder. After
1 min, the cylinder was removed and the observations
started. The antennae of the target ant were abscised
prior to the test, so that the target ant was alive but
passive. This allowed us to record the behaviour of the
focal ant against a live ant, without the risk of the target
ant interfering. The following items were recorded during
3 min: (i) no contact, (ii) antennation, (iii) mandible
opening, (iv) biting and (v) gaster flexing. For each
aggression test, an aggression index was calculated
according to Errard & Hefetz (1997). In addition, avoid-
ance behaviour was calculated, as defined by the dura-
tion of ‘no contact’ after the first contact, divided by the
total time after the first contact. Before day 3, five of
the treated ants died, making a total of 55 aggression
tests conducted.
Statistical analyses
Analyses of all data were carried out in R version 2.8.1(R
Development Core Team, 2008). Generalized Linear
Mixed Models (GLMM) were used to investigate the
effect of treatment, colony, and time on the duration spent
by focal ants (treated and controls) in a given location and
on the duration spent performing a given behaviour
(trophallaxis, antennation, self-grooming, being
groomed, grooming other). A GLMM with a binomial
error structure and logit link function was used to
examine the effect of treatment, time and colony on the
likelihood that focal ants were observed in a given location
and engaged in a given behaviour. Finally, locations and
behaviours frequencies were analysed using GLMM with
Poisson errors. For all GLMM analyses, we used the lmer
function of the lme4 library. Time was treated as a
continuous variable, and colony and treatment were
treated as categorical variables. Full models included
treatment, colony and time as explanatory variables,
and all relevant second-order interactions. Time and
treatment were specified as fixed effects and colony as a
random effect. We also included ant identity as a nested
random factor within colony to account for the repeated
behavioural measurements performed on individual ants
(i.e. pseudoreplications) (Pinheiro & Bates, 2000).
From the full models, we removed nonsignificant
(P> 0.05) terms, starting with the least significant
second-order interaction. Only terms for which removal
significantly reduced the explanatory power of the model
were retained in the minimal model (Crawley, 2007).
Model selection was based upon the likelihood ratio test
(LRT) statistics. Terms significance was also estimated
with an alternative method by Markov chain Monte
Carlo sampling (using the function mcmcsamp from
lme4 (Baayen et al., 2008)). The two approaches yielded
the same results, and only the LRT tests are reported on.
Peak areas of the integrated cuticular hydrocarbon
profiles were normalized according to Aitchison (1986)
and used as variables in a principal component analysis,
followed by
, with both treatment and colony as
categorical predictors. Effect size was calculated using the
software G
3.1.3 (Erdfelder et al., 1996).
Results from the aggression tests were square root
transformed, and afterwards analysed using a general
linear model, with both treatment and colony as cate-
gorical predictors.
Experiment 1: Behavioural study
All data presented are based on a total of 127 observation
periods of 240 s for treated ants and 248 observation
periods of 240 s for the control ants.
The survival of infected ants was much lower than
control, uninfected ants (Fig. 1). Although all infected ants
died by day 7 post-exposure, none of the control ants died
over the course of the period of survival observation
(7 days). We found no effect of treatment on antennation
time (LRT v
= 0.35, P= 0.55, Fig. 2a, see Table S1 for full
model details). However, antennation significantly varied
among ant colonies (LRT v
= 12.77, P= 0.01) and
decreased over the course of the experiment (LRT
= 19.77, P< 0.001, Fig. 2a). Treated ants spent signif-
icantly less time performing trophallaxis than their unin-
fected counterparts (mean ± SE; 8 ± 0.7 s and 16.7 ± 1 s
respectively; LRT v
= 4.9, P= 0.02, Fig. 2b). We also
found that trophallaxis decreased over time in both treated
and control workers (LRT v
= 4.7, P= 0.03, Fig. 2b).
Treated ants were groomed significantly more often by
nestmates than control ants (10.4 ± 2.90 s and
Time (days)
Infected ants
Control ants
Fig. 1 Survivorship of infected and uninfected individuals. Propor-
tion of infected and control ants that survived during the first 7 days
of observation.
Sick ants become unsociable 345
ª2011 THE AUTHORS. J. EVOL. BIOL. 25 (2012) 342–351
3.98 ± 0.99 s for treated and control individuals respec-
tively, LRT v
=8,P= 0.005, Fig. 2c), although this only
occurred on days 1 and 2 post-treatment (time by treat-
ment interaction: LRT v
= 10.1, P= 0.02, Fig. 2c). We
also found a significant colony main effect with individuals
from colonies denoted 1 and 3 spending more time being
groomed than individuals from other colonies (LRT
= 9.7, P= 0.04). There was a significant colony-by-
treatment interaction indicating that the intensity of care
directed towards treated individuals varied among colonies
(LRT v
= 10.3, P= 0.04). Treated individuals almost
never performed allo-grooming towards nestmates from
day 1 post-treatment until death, although the difference
with control workers was not statistically significant
(1 ± 0.09 s and 3.2 ± 0.2 s for treated and control indi-
viduals respectively, LRT v
= 2.2, P= 0.13, Fig. 2d).
There was a strong effect of infection on self-grooming,
with treated workers spending significantly more time
performing this activity than control workers (26 ± 2.3 s
and 15 ± 1 s respectively, LRT v
= 11.8, P< 0.001,
Fig. 2e). We also found a significant effect of time on self-
grooming with both treated and control ants self-grooming
less and less over the course of the experiment (LRT v
P= 0.04, Fig. 2e). The observed decrease in self-grooming
over time was much steeper in treated individuals than in
uninfected counterparts (time by treatment interaction:
=7, P= 0.008, Fig. 2e). Finally, self-grooming
significantly varied among colonies with individuals from
colony 4 spending less time self-grooming than individuals
from other colonies (LRT v
= 11, P= 0.03).
The probability of becoming infected upon pathogen
inoculation was not influenced by self-grooming and or
(a) (b)
Fig. 2 Behaviours of infected and uninfected ants overtime. Infected ants did not differ from control ants in antennation time. They
performed significantly less trophallaxis and less allo-grooming (nonsignificant trend). However, infected ants were groomed significantly more
by their nestmates, as well as increasing the time spent self-grooming. Data are presented as mean ± SE.
346 N. BOS ET AL.
ª2011 THE AUTHORS. J. EVOL. BIOL. 25 (2012) 342–351
on receiving allo-grooming. In other words, the four
individual ants that remained uninfected upon treatment
with the pathogen were not those that spent most time
self-grooming or being groomed (GLM binomial, self-
grooming (Odds Ratio = 1.2, 95% confident interval
(CI) = (0.9,1.5), P= 0.3, being groomed OR = 1.3,
CI = (0.8,1.8), P= 0.5).
We found that treated ants spent significantly less time
in the nest than control ants (treated ants: 170.6 ± 14.9 s;
control ants 235.3 ± 15.1 s; GLMM, v
= 4.13, P= 0.04,
Fig. 3a). Control workers indeed spent almost all their
time in the nest whereas infected individuals spent an
increasing amount of time in the foraging arena as the
infection progressed (treatment by time interaction: LRT
= 30.6, P< 0.001, Fig. 3a). There was also a significant
effect of treatment on the time spent with the brood.
Treated individuals had almost no interaction with the
brood from day 1 post-infection until death (treated ants:
2.5 ± 0.23 s; control ants: 19.1 ± 1.2 s; LRT v
= 3.5,
P= 0.04, Fig. 3b). Control ants spent 8.13% ± 1.45 of
the time they spent in the nest with the brood. In contrast,
treated ants spent only 1.39% ± 0.99 of their time inside
the nest with the brood.
Overall, the analyses of location and behaviour fre-
quency and likelihood yielded similar conclusions (see
Table S1 for full model details of frequency and likeli-
hood analyses). In other words, if treated ants spent more
time performing a given behaviour than control ants,
they were also more likely to be engaged in the
behaviour (likelihood analysis) and performed it more
often (frequency analysis).
Experiment 2: Chemical analysis
The principal component analysis based on the normal-
ized area of the 36 identified peaks extracted six principal
components explaining 93.9% of the observed variation.
A plot of the first three PCs (explaining the 79.5% of the
total variance) shows that infected and uninfected ants
are grouping together within each subcolony (Fig. 4). No
significant difference was found between treated and
control ants (
, Pillai V = 0.36, F
= 2.11,
P= 0.09, effect size f
V = 0.54), indicating that the
chemical profile did not change significantly after infec-
tion (Fig. 4, Fig. S1 and Table S2). Colonies were signif-
icantly different from each other (
, Pillai
V = 1.98, F
= 318.23, P< 0.01); however, no signif-
icant interaction between colony and infection was
found (
, colony x treatment, Pillai V = 0.59,
= 1.68, P= 0.10).
(a) (b)
Fig. 3 Changes in space use of infected and uninfected ants overtime. Infected ants spend less time than control ants both in
the nest (a) and with the brood (b) over the course of the experiment. Data are presented as mean ± SE.
Fig. 4 Chemical analysis of control and infected ants. Scatterplot of
the first three factors of the PCA (explaining 79.5% of the total
variation, Factor 1: 48.0%; Factor 2: 20.9%; Factor 3: 10.6%),
showing that treated ants were not separated from control ants.
Symbols denote different colonies. Black symbols are infected ants,
and white symbols are control ants.
Sick ants become unsociable 347
ª2011 THE AUTHORS. J. EVOL. BIOL. 25 (2012) 342–351
Experiment 3: Aggression tests
Treated ants, tested on day 3 post-inoculation, showed a
higher level of aggression against non-nestmates than
uninfected control ants (GLM, F= 6.36, P= 0.01,
Fig. 5a). There was no significant effect of colony on
aggression level (GLM, F= 1.84, P=0.17); however, a
marginally significant effect on aggression was found in
the interaction between infected x colony (indicating
that the effect of infection on aggression was not
consistent across colonies, GLM, F= 3.26, P= 0.047).
Also, treated ants showed less avoidance behaviour when
compared to control individuals (GLM, F= 4.36,
P=0.04, Fig. 5b). No significant effect of colony was
found on avoidance (GLM, F= 1.80, P= 0.18); however,
a nonsignificant trend was found in the interaction
between infected x colony (GLM, F= 3.15, P=0.052.
The main aim of this study was to investigate how
treatment with a pathogen affects behaviour and com-
munication in ants. We found that treatment with the
generalist entomopathogenic fungus M. brunneum pro-
foundly changes the behaviour of both infected ants and
their nestmates.
Treated ants spent significantly less time performing
trophallaxis than their uninfected counterparts. We also
found that the amount of trophallaxis decreased over the
course of the experiment in both treated and control
workers. A possible explanation for this is that almost no
foraging activity was observed in both groups, suggesting
that ants were satiated and thus newly provided food was
not brought into the nest, decreasing the need for
trophallaxis. Our results are different from those of De
Souza et al. (2008) and Hamilton et al. (2011), who found
an increase in trophallaxis after an immune challenge,
but in these studies ants were either socially isolated (for
24 or 8 h, De Souza et al., 2008) or the receiver was
starved (Hamilton et al., 2011), thus generally increasing
the need for trophallaxis. Moreover, in our case the ants
were exposed to a real parasite, whereas in the above
cited studies the ants were immunized and could share
their immunological factors with nestmates without the
risk of spreading the disease; however, whether the ants
can distinguish between an immune challenge and an
infection that will lead to its death remains to be
We found that treated ants exhibited an increase in
self-grooming and received more allo-grooming in com-
parison with control nestmates on days 1 and 2 (i.e.
Rosengaus et al., 1998; Walker & Hughes, 2009).We
interpret this behaviour as a way to get rid of the conidia
(the infective units of the parasite) present on the cuticle
(Walker & Hughes, 2009). Indeed, salivary secretions
with antimicrobial properties (e.g. Bulmer et al., 2009)
could be spread through allo-grooming and thus poten-
tially decrease the microbial load. However, in this study
we did not find that the probability of becoming infected
upon inoculation with the pathogen depended on self-
grooming and or on receiving allo-grooming, possibly
because of the high concentration of conidia used.
Another possible function of grooming treated individu-
als could be for the uninfected nestmates to gain
resistance towards the pathogen (Ugelvig & Cremer,
2007; Walker & Hughes, 2009).
The fact that in the first 2 days the treated ants
received more grooming than control ants suggests that
nestmates can detect whether an ant is in need of
grooming. Several nonmutually exclusive explanations
(a) (b)
Fig. 5 Aggression tests. Aggression index (a) and avoidance index (b) of infected and control ants, measured on day 3 post-exposure
to Metarhizium brunneum. Infected ants were more aggressive and displayed less avoidance behaviour than control ants.
348 N. BOS ET AL.
ª2011 THE AUTHORS. J. EVOL. BIOL. 25 (2012) 342–351
may account for this observation. First, infected ants may
display an altered cuticular profile (as shown in immune
challenged Lasius niger queens, Holman et al., 2010),
inducing nestmates to groom them; however, we found
no evidence for this using GC-MS, as infection did not
affect the cuticular chemical profile; however, we con-
ducted chemical analysis on infected ants, and not on
infectious ants, so a possibility of ants changing their
profile after exposure in the first 2 days exists. Second,
ants might detect that another ant is grooming itself and
react by starting to groom this ant. We also did not find
support for this second hypothesis. Ants groomed them-
selves for a total of 816 times, and in only 17 of these
cases (2.1%), self-grooming was followed by being
groomed. Third, an ant may show other behavioural
abnormalities (e.g. vibration signals), which are difficult
for the observer to notice and the presence of conidia
might be detected by conspecifics using other cues (e.g.
tactile, visual). This hypothesis remains to be tested.
From day three on, no significant difference in groom-
ing behaviour was found between treated and control
individuals. However, at this point in time, striking
differences started to appear in other behaviours, which
is in accordance with the study of Gillespie et al. (2000),
showing that penetration of the conidia through the
cuticle (thus, actually infecting the ant) takes 2–4 days.
After penetration, removal of conidia will have no effect,
and only the insect’s physiological immune system can
save it, which might explain the decrease in self- and allo-
grooming after day two. Treated ants became less sociable
and spent more time outside the nest. Also, even when
inside the nest, the ants spent less time in contact with the
brood. Spending less time in the nest and with the brood
might be a way to increase the infected individual’s
inclusive fitness, by making sure related individuals are
not exposed to the pathogen. This altruistic self-removal
has recently been found in Temnothorax ants (Heinze &
Walter, 2010) and honey bees (Rueppell et al., 2010).
We found that treatment with the pathogen did not
alter the cuticular hydrocarbon profile of ants, suggesting
that they do not signal their physiological status with a
change in recognition cues. However, the results show
a nonsignificant trend, suggesting that infected indivi-
duals have a slightly different chemical profile than that
of uninfected individuals. The reason behind this slight
difference is currently unclear. Besides a direct effect of
infection on ants cuticular profile, we cannot rule out the
existence of an indirect effect through nest desertion, as
isolation has been shown to change the hydrocarbon
profile of ants (Boulay et al., 2000; Bos et al., 2011). In
our experiment, this self-isolation of infected individuals
might thus have affected their cuticular hydrocarbon
profile. We also cannot exclude the use of more volatile
compounds, which we did not investigate. However, the
fact that infected ants spend more time outside the nest,
and especially die outside, possibly makes signalling their
physiological state unnecessary.
We found for the first time that infected ants increase
their level of aggression towards non-nestmates. The
overall level of aggression was higher, and infected ants
expressed less avoidance behaviour than their uninfected
counterparts. We interpret this result as if the treated,
infected ant has ‘nothing to lose’. The ant will face
certain death, and when confronted with the choice of
deterring or avoiding possible threats to the colony, it will
act aggressively, instead of displaying the more usual
avoidance behaviour. The observed behaviours (self-
removal and increased aggression) are reminiscent of
what is also found in humans, where sick individuals
become reclusive and irritable, isolating themselves from
other individuals (Loehle, 1995).
In ants, age-dependant polyethism exists (O’donnell &
Jeanne, 1995), where older workers will take on more
risky tasks, such as foraging (Woyciechowski & Kozlow-
ski, 1998; Woyciechowski & Moron, 2009). Foraging
increases the time spent outside of the nest. Also, recent
evidence suggests that foragers are more aggressive
against non-nestmates than nurses (J. Larsen, unpub-
lished). It is possible that infection triggers the same
pathways that are involved in the process of ageing (e.g.
HIV in humans, Ances et al., 2010), and as a result
infected ants spend more time outside of the nest and
become more aggressive. One way in which both age and
health status could elicit the same behavioural changes in
ants, is if an ant cannot assess its age, but only its
longevity, which is affected both by age and by health
status (Tofilski, 2002).
Our results emphasize that parasite pressure has
shaped many behaviours in the evolution of social
insects. In particular, unsociable behaviours can be
selected for in social insects when benefits of preventing
disease outbreaks in the nest outweigh the potential cost
of losing workers (Rueppell et al., 2010).
It is easy to imagine how this is true in the case of
infection, where one infected individual could poten-
tially infect its entire colony, destroying it in the process.
We would like to thank all members of the Centre for
Social Evolution for providing a pleasant working envi-
ronment. Many thanks to Louise Lee Munck Larsen for
helping with rearing M. brunneum and to Jelle van
Zweden for comments and discussion. This work was
supported by the Danish National Research Foundation
and by a grant from the Faculty of Science, University of
Copenhagen to PdE.
Aitchison, J. 1986. The Statistical Analysis of Compositional Data.
The Blackburn Press, Caldwell.
Ances, B.M., Vaida, F., Yeh, M.J., Liang, C.L., Buxton, R.B.,
Letendre, S. et al. 2010. HIV infection and aging independently
Sick ants become unsociable 349
ª2011 THE AUTHORS. J. EVOL. BIOL. 25 (2012) 342–351
affect brain function as measured by functional magnetic
resonance imaging. J. Infect. Dis. 201: 336–340.
Andersen, S.B., Gerritsma, S., Yusah, K.M., Mayntz, D., Hywel-
Jones, N.L., Billen, J. et al. 2009. The life of a dead ant: the
expression of an adaptive extended phenotype. Am. Nat. 174:
Baayen, R.H., Davidson, D.J. & Bates, D.M. 2008. Mixed-effects
modeling with crossed random effects for subjects and items.
J. Memory Lang. 59: 390–412.
Bischoff, J.F., Rehner, S.A. & Humber, R.A. 2009. A multilocus
phylogeny of the Metarhizium anisopliae lineage. Mycologia
101: 512–530.
Boomsma, J.J., Schmid-Hempel, P., Hughes, W.O.H., Fellowes,
M.D.E., Holloway, G.J. & Rolff, J. 2005. Life histories and
parasite pressure across the major groups of social insects.
Insect Evolutionary Ecology. Proceedings of the Royal Entomological
Society’s 22nd Symposium. 139–175.
Bos, N., Grinsted, L. & Holman, L. 2011. Wax on, wax off: nest
soil facilitates indirect transfer of recognition cues between ant
nestmates. PLoS ONE 6: e19435.
Boulay, R., Hefetz, A., Soroker, V. & Lenoir, A. 2000. Camponotus
fellah colony integration: worker individuality necessitates
frequent hydrocarbons exchanges. Anim. Behav. 59: 1127–1133.
Bourke, A.F.G. 2011. Principles of Social Evolution. Oxford
University Press, Oxford.
Bull, J.J., Molineux, I.J. & Rice, W.R. 1991. Selection of
benevolence in a host-parasite system. Evolution 45: 875–882.
Bulmer, M.S., Bachelet, I., Raman, R., Rosengaus, R.B. &
Sasisekharan, R. 2009. Targeting an antimicrobial effector
function in insect immunity as a pest control strategy. Proc.
Natl Acad. Sci. USA 106: 12652–12657.
Calleri, D.V., Rosengaus, R.B. & Traniello, J.F.A. 2006. Disease
and colony establishment in the dampwood termite Zooterm-
opsis angusticollis: survival and fitness consequences of infec-
tion in primary reproductives. Insect. Soc. 53: 204–211.
Castella, G., Chapuisat, M. & Christe, P. 2008. Prophylaxis with
resin in wood ants. Anim. Behav. 75: 1591–1596.
Chapuisat, M., Oppliger, A., Magliano, P. & Christe, P. 2007.
Wood ants use resin to protect themselves against pathogens.
Proc. R. Soc. B 274: 2013–2017.
Christe, P., Oppliger, A., Bancala
`, F., Castella, G. & Chapuisat,
M. 2003. Evidence for collective medication in ants. Ecol. Lett.
6: 19–22.
Cotter, S.C. & Kilner, R.M. 2010. Personal immunity versus
social immunity. Behav. Ecol. 21: 663–668.
Crawley, M.J. 2007. The R Book. Wiley, Hoboken, New Jersey.
Cremer, S. & Sixt, M. 2009. Analogies in the evolution of
individual and social immunity. Philos. Trans. R. Soc. Lond. B
Biol. Sci. 364: 129–142.
Cremer, S., Armitage, S.A.O. & Schmid-Hempel, P. 2007. Social
immunity. Curr. Biol. 17: R693–R702.
De Souza, D.J., Van Vlaenderen, J., Moret, Y. & Lenoir, A. 2008.
Immune response affects ant trophallactic behaviour. J. Insect
Physiol. 54: 828–832.
Erdfelder, E., Faul, F. & Buchner, A. 1996. GPOWER: A general
power analysis program. Behav. Res. Methods Instrum. Comput.
28: 1–11.
Errard, C. & Hefetz, A. 1997. Label familiarity and discriminatory
ability of ants reared in mixed groups. Insect. Soc. 44: 189–198.
Gillespie, J.P., Bailey, A.M., Cobb, B. & Vilcinskas, A. 2000.
Fungi as elicitors of insect immune responses. Arch. Insect
Biochem. Physiol. 44: 49–68.
Hamilton, C., Lejeune, B.T. & Rosengaus, R.B. 2011. Trophal-
laxis and prophylaxis: social immunity in the carpenter ant
Camponotus pennsylvanicus. Biol. Lett. 7: 89–92.
Heinze, J. & Walter, B. 2010. Moribund ants leave their nests to
die in social isolation. Curr. Biol. 20: 249–252.
Holman, L., Jorgensen, C.G., Nielsen, J. & d’Ettorre, P. 2010.
Identification of an ant queen pheromone regulating worker
sterility. Proc. R. Soc. B 277: 3793–3800.
Hughes, W.O.H. & Boomsma, J.J. 2004. Genetic diversity and
disease resistance in leaf-cutting ant societies. Evolution 58:
Hughes, W.O.H., Eilenberg, J. & Boomsma, J.J. 2002. Trade-offs
in group living: transmission and disease resistance in leaf-
cutting ants. Proc. R. Soc. B 269: 1811–1819.
Lacey, L.A. & Brooks, W.M. 1997. Initial handling and diagnosis
of diseased insects. In: Manual of Techniques in Insect Pathology
(L.A. Lacey, ed.), pp. 1–16. Academic, London.
Lefevre, T., Lebarbenchon, C., Gauthier-Clerc, M., Misse, D.,
Poulin, R. & Thomas, F. 2009. The ecological significance of
manipulative parasites. Trends Ecol. Evol. 24: 41–48.
Lenoir, A., Depickere, S., Devers, S., Christides, J.P. & Detrain, C.
2009. Hydrocarbons in the Ant Lasius niger: from the cuticle to
the nest and home range marking. J. Chem. Ecol. 35: 913–921.
Loehle, C. 1995. Social barriers to pathogen transmission in wild
animal population. Ecology 76: 326–335.
Maynard Smith, J. & Szathmary, E. 1995. The Major Transitions in
Evolution. Oxford University Press, Oxford.
McCallum, H., Barlow, N. & Hone, J. 2001. How should
pathogen transmission be modelled? Trends Ecol. Evol. 16:
Moore, J. 2002. Parasites and the Behavior of Animals. Oxford
University Press, Oxford.
Myers, J.H. & Rothman, L.E. 1995. Virulence and transmitssion
of infectious diseases in humans and insects- Evolutionary and
demographic patterns. Trends Ecol. Evol. 10: 194–198.
O’donnell, S. & Jeanne, R.L. 1995. Implications of senescence
patterns for the evolution of age polyethism in eusocial
insects. Behav. Ecol. 6: 269–273.
Oi, D.H. & Pereira, R.M. 1993. Ant behavior and microbial
pathogens (Hymenoptera, Formicidae). Fla. Entomol. 76: 63–74.
Ottoni, E.B. 2000. Etholog 2.2 - a tool for the transcription and
timing of behavior observation sessions. Behav. Res. Methods
Instrum. Comput. 32: 446–449.
Pinheiro, J.C. & Bates, D.M. 2000. Mixed-Effects Models in S and
S-plus. Springer-Verlag, New York.
R Development Core Team 2008. R: A Language and Environment
for Statistical Computing. R Foundation for Statistical Comput-
ing, Vienna.
Rosengaus, R.B., Maxmen, A.B., Coates, L.E. & Traniello, J.F.A.
1998. Disease resistance: a benefit of sociality in the
dampwood termite Zootermopsis angusticollis (Isoptera : Term-
opsidae). Behav. Ecol. Sociobiol. 44: 125–134.
Rueppell, O., Hayworth, M.K. & Ross, N.P. 2010. Altruistic self-
removal of health-compromised honey bee workers from their
hive. J. Evol. Biol. 23: 1538–1546.
Schmid-Hempel, P. 1998. Parasites in Social Insects. Princeton
University Press, Princeton, New Jersey.
Simone, M., Evans, J.D. & Spivak, M. 2009. Resin collection and
social immunity in Honey bees. Evolution 63: 3016–3022.
Starks, P.T., Blackie, C.A. & Seeley, T.D. 2000. Fever in
honeybee colonies. Naturwissenschaften 87: 229–231.
350 N. BOS ET AL.
ª2011 THE AUTHORS. J. EVOL. BIOL. 25 (2012) 342–351
Tofilski, A. 2002. Influence of age polyethism on longevity of
workers in social insects. Behav. Ecol. Sociobiol. 51: 234–237.
Traniello, J.F.A., Rosengaus, R.B. & Savoie, K. 2002. The
development of immunity in a social insect: evidence for
the group facilitation of disease resistance. Proc. Natl Acad. Sci.
USA 99: 6838–6842.
Ugelvig, L.V. & Cremer, S. 2007. Social prophylaxis: group
interaction promotes collective immunity in ant colonies.
Curr. Biol. 17: 1967–1971.
Van Zweden, J.S., Dreier, S. & d’Ettorre, P. 2009. Disentangling
environmental and heritable nestamte recognition cues in a
carpenter ant. J. Insect Physiol. 55: 159–164.
Walker, T.N. & Hughes, W.O.H. 2009. Adaptive social immunity
in leaf-cutting ants. Biol. Lett. 5: 446–448.
Wilson-Rich, N., Stuart, R.J. & Rosengaus, R.B. 2007. Suscep-
tibility and behavioral responses of the dampwood termite
Zootermopsis angusticollis to the entomopathogenic nema-
tode Steinernema carpocapsae. J. Invertebr. Pathol. 95: 17–25.
Wilson-Rich, N., Spivak, M., Fefferman, N.H. & Starks, P.T.
2009. Genetic, individual, and group facilitation of disease
resistance in insect societies. Annu. Rev. Entomol. 54: 405–423.
Woyciechowski, M. & Kozlowski, J. 1998. Division of labor by
division of risk according to worker life expectancy in the
honey bee (Apis mellifera L.). Apidologie 29: 191–205.
Woyciechowski, M. & Moron, D. 2009. Life expectancy and
onset of foraging in the honeybee (Apis mellifera). Insect. Soc.
56: 193–201.
Yanagawa, A. & Shimizu, S. 2007. Resistance of the termite,
Coptotermes formosanus Shiraki to Metarhizium anisopliae due to
grooming. Biocontrol 52: 75–85.
Yanagawa, A., Yokohari, F. & Shimizu, S. 2008. Defense
mechanism of the termite, Coptotermes formosanus Shiraki, to
entomopathogenic fungi. J. Invertebr. Pathol. 97: 165–170.
Yek, S.H. & Mueller, U.G. 2010. The metapleural gland of ants.
Biol. Rev. 86: 774–791.
van Zweden, J.S. & d’Ettorre, P. 2010. Nestmate recognition in
social insects and the role of hydrocarbons. In: Insect Hydro-
carbons, (G.J. Blomquist & A.G. Bagneres, eds), pp. 222–243.
Cambridge University Press, Cambridge.
Supporting information
Additional Supporting Information may be found in the
online version of this article:
Table S1 Changes in behaviour and space use after
Table S2 Chemical analysis of control and infected ants.
Figure S1 Chemical analysis of control and infected ants.
As a service to our authors and readers, this journal
provides supporting information supplied by the authors.
Such materials are peer-reviewed and may be reorga-
nized for online delivery, but are not copy-edited or
typeset. Technical support issues arising from supporting
information (other than missing files) should be
addressed to the authors.
Received 25 July 2011; revised 25 October 2011; accepted 26 October
Sick ants become unsociable 351
ª2011 THE AUTHORS. J. EVOL. BIOL. 25 (2012) 342–351

Supplementary resource (1)

... Infection avoidance is the first line of behavioural defence, where hosts modify their behaviour if they perceive an infection risk in their environment or from conspecifics [8][9][10][11]. This may include spatial or habitat avoidance [12,13], trophic avoidance [11,14,15] and social avoidance [11,16]. Nevertheless, it is rarely possible to completely avoid infection, as many common infection routes involve activities that are central to organismal physiology and fitness, including foraging and feeding. ...
... Social avoidance of infection is a widespread mechanism of defence in the animal kingdom [9,34]. Sick individuals may decrease social connectivity due to lethargic behaviour or actively self-isolate [16,35], but they can also be avoided by healthy individuals to avoid direct routes of infection [9,20,36]. This social behavioural flexibility leads to detectable changes in the group social structure, which affects the risk of contagion among individuals [2,37]. ...
Full-text available
Identifying how infection modifies host behaviours that determine social contact networks is important for understanding heterogeneity in infectious disease dynamics. Here, we investigate whether group social behaviour is modified during bacterial infection in fruit flies ( Drosophila melanogaster ) according to pathogen species, infectious dose, host genetic background and sex. In one experiment, we find that systemic infection with four different bacterial species results in a reduction in the mean pairwise distance within infected female flies, and that the extent of this change depends on pathogen species. However, susceptible flies did not show any evidence of avoidance in the presence of infected flies. In a separate experiment, we observed genetic- and sex-based variation in social aggregation within infected, same-sex groups, with infected female flies aggregating more closely than infected males. In general, our results confirm that bacterial infection induces changes in fruit fly behaviour across a range of pathogen species, but also highlight that these effects vary between fly genetic backgrounds and can be sex-specific. We discuss possible explanations for sex differences in social aggregation and their consequences for individual variation in pathogen transmission.
... Holding stabilizes the pellets, reducing movement and tension for cutting or other processing behaviors and the energy expenditure (Garrett et al. 2016). The pellet fragmentation behavior reduces this substrate into tiny parts (Garrett et al. 2016) making it easier for smaller ant workers to perform other processing behaviors (Bos et al. 2011). Thus, there is a possibility that these behaviors provide direct contact among workers and the insecticide in toxic baits. ...
... Allogrooming performed by workers and associated with the recognition of nestmates is triggered when a substance is recognized in the nestmate tegument (Soroker et al. 1995;Hughes et al. 2002). This also occurs with self-grooming, because ants increase this behavior in response to the exposure to harmful substances (Bos et al. 2011;Theis et al. 2015;Camargo et al. 2017b). ...
Toxic baits, containing the active ingredients sulfluramid or fipronil, are the main methods to control leaf-cutting ants of the genera Atta Fabricius, 1805, and Acromyrmex Mayr, 1865 (Hymenoptera: Formicidae). However, the insecticide dispersion among members of the colony during the control needs further studies. We studied whether the behaviors of allogrooming, self-grooming, and contact among individuals spread the insecticides among those of the colony. The insecticides sulflura-mid and fipronil (0.1% and 1.0% (w/w)) were applied topically in groups of workers of Atta sexdens (Linnaeus, 1758), and the social interactions among them with or without insecticide were studied. In addition, toxic baits (sulfluramid or fipronil) were provided to colonies and their behavioral acts were observed. At the end of the experiment, colony mortality, number and mass of dead workers, and mass of wet waste were compared between ant nests receiving baits and ants with topical application. In the topical application, behavioral analysis showed higher interaction between ants in the colonies and touch and allogrooming behaviors as the most frequent in those that received the concentrations of sulfluramid. In the baits, the behavior of licking the pellet and allogrooming was more frequent. Colony mortality was faster for those that received topical application, especially with the insecticide fipronil (0.1%). However, the number and mass of dead workers was similar between topical application and toxic baits. In the toxic baits, the licking behavior of the bait pellets and subsequent allo-grooming probably dispersed the insecticides. In the topical application, the route of the insecticide occurred by excessive touches among workers, with subsequent allogrooming. Thus, allogrooming, self-grooming, and touching among workers increased the dispersion of insecticides among members of the ant colonies.
... This strong detection ability allows social insects to adjust their behavior to reduce disease transmission risk. In several species of ants and bees, infected and contagious individuals self-isolate themselves, either by reducing their contact with nestmates or by leaving the nest entirely (Walker and Hughes, 2009;Chapuisat, 2010;Heinze and Walter, 2010;Bos et al., 2012;Stroeymeyt et al., 2018;Geffre et al., 2020;Alciatore et al., 2021). ...
Full-text available
Social insect colonies are characterized by an efficient division of labor, allowing high-value individuals (i.e., reproductives and brood) to be sheltered from tasks associated with increased risk of pathogen exposure, such as foraging or corpse disposal. This social organization helps limit the transmission of disease throughout the colony. Further, individuals can actively respond to imminent disease threats by altering their behaviors as a means of social immunity. In subterranean termites, although workers typically avoid detected pathogens, they can be attracted to pathogen cues when a nestmate is infected. Infected termites are usually groomed, but they may instead be cannibalized if the infection has already become lethal. The mechanisms governing these changes in behavior are unclear. We set out to examine immediate changes in individual behaviors, investigating the role that the infected individual plays in communicating its infection status to nestmates. We also assessed gradual changes in social organization after the re-introduction of an infected termite to the colony. Our results reveal that infected termites likely do not signal their infection status to nestmates through shaking behaviors and reduced movements, suggesting the occurrence of other mechanisms used in communicating infection. We also found that infected termites do not self-isolate and may travel to the densest part of the colony, where they can potentially benefit from grooming by large groups of nestmates. These results provide new insights into how individual changes in immune behaviors contribute to overall colony health, highlighting that, at early stages of infection, termites favor a rescuing strategy rather than isolation and/or cannibalization.
... A major criticism of M. anisopliae is the relatively slow kill speed in field conditions [9]. Although M. anisopliae is a good biological control agent against locusts, locusts show resistance to foreign substances through their immune systems, such Agriculture 2022, 12, 538 2 of 9 as via thermoregulation and innate immunity [10,11]. Therefore, it may be beneficial to enhance the virulence of M. anisopliae to ensure enough virulence to kill locusts after initial feeding [12]. ...
Full-text available
RNA interference (RNAi) induces gene silencing in order to participate in immune-related pathways. Previous studies have shown that the LmSerpin1 gene upregulates the immune responses of Locusta migratoria manilensis, and that the infection of locusts with Metarhizium anisopliae can be enhanced through the injection of the interfering dsSerpin1. In this study, dsRNA was synthesized from the Serpin1 gene and fed to the third instars of L. m. manilensis at concentrations of 5 μg, 10 μg, and 20 μg. Feeding dsSerpin1 increased the mortality of L. m. manilensis, and 20 μg dsSerpin1 had the highest lethality, followed by 10 μg dsSerpin1, with the lowest being observed at 5 μg dsSerpin1. Serpin1 silencing and temperature fever induction by feeding with dsSerpin1 had a strong synergistic effect on M. anisopliae, showing a dose-dependent response. When L. m. manilensis were fed on a diet containing 20 μg dsRNA and M. anisopliae (2.5 × 108 spores/g wheat bran), 3-day mortalities significantly increased. The dsSerpin1 plus M. anisopliae treatments resulted in a 6-day mortality of 60%, accelerating insect death. These results indicated that feeding with dsSerpin1 could be an effective way to control pests and that the control effect and dsRNA concentration were dose-dependent within the first seven days. In particular, 20 μg was an effective dosage to enhance the insecticidal efficiency of M. anisopliae.
... Motivated by past findings that immune-challenged ants (Bos et al 2012) and bees Kazlauskas et al 2016) engage in fewer social interactions, experiment 3 tested whether pairs of bees in which one member had received an immune challenge spent less time in close contact than control pairs. We recorded a statistically significant decline in the proportion of time spent in contact, suggesting a behavioural response to immune challenge in the treated individual, the healthy individual paired with them, or both. ...
Full-text available
Animals living in large colonies are especially vulnerable to infectious pathogens and may therefore have evolved additional defences. Eusocial insects supplement their physiological immune systems with ‘social immunity’, a set of adaptations that impedes the entrance, establishment, and spread of pathogens in the colony. We here find that honey bee workers ( Apis mellifera ) that had been experimentally immune-challenged with bacterial lipopolysaccharide (LPS) often exited the hive and subsequently died; some individuals were dragged out by other workers, while others appeared to leave voluntarily. In a second experiment, we found that healthy workers treated with surface chemicals from LPS-treated bees were evicted from the hive more often than controls, indicating that immune-challenged bees produce chemical cues or signals that elicit their eviction. Thirdly, we observed pairs of bees under lab conditions, and found that pairs spent more time apart when one member of the pair had received LPS, relative to controls. Our findings suggest that immune-challenged bees altruistically banish themselves, and that workers evict sick individuals which they identify using olfactory cues, putatively because of (kin) selection to limit the spread of pathogens within colonies. Significance statement Just as in humans, animals living in large groups must contend with infectious diseases. Social insects such as honey bees have evolved a range of behavioural and organisational defences against disease, collectively termed ‘social immunity’. Here, we describe experiments in which we introduced immune-stimulated bee workers into hives to study social immunity. We find that bees that were wounded or immune-challenged were more likely to leave the hive—resulting in their death—compared to healthy controls. Some of the bees leaving the hive were ejected by other workers, while some left the hive seemingly by choice: we thus find evidence for both ‘banishment’ of immune-challenged bees and self-imposed exile. Furthermore, using experiments transferring chemical signals between healthy and immune stimulated bees, we establish that the latter are identified for banishment by the chemicals present on their body surface.
... avoiding interactions with O.F, potentially carrying more germs, in accordance with the fact that they have been shown in this study to have fewer antibiotic metabolites. It has also been shown that sick ants isolate themselves before dying (Heinze and Walter 2010;Bos et al. 2012). Both cases would lead to the social isolation of this group and therefore to a weaker synthesis of communication molecules. ...
Full-text available
My thesis analyses how the social environment influences longevity in various animal species (vertebrates and invertebrates) in a gradient of social complexity (gregariousness, cooperative breeding, eusociality). Thanks to complementary measurements (behavioural observations, oxidative stress, telomere length, proteomics, metabolomics), my work reflects contrasting ageing mechanisms depending on the species studied. For example, in birds, longer telomeres are associated with a favourable social environment and greater longevity. On the contrary, among the 10 species of ants studied, the longest-lived ants had shorter telomeres. Furthermore, the complete and combined study of the proteome and metabolome of ants has highlighted certain mechanisms (e.g., sirtuins, mTOR, anti-cancer mechanisms), the universality of which opens the way to a better understanding of the harmful effect of social stress on human health.
Insect behavior is controlled by genes that specify transcription factors and hormones that affect how the nervous system responds to external stimuli and physiological state. Internal clocks residing in many cells of several organs are responsible for circadian rhythms and photoperiodism. Insects are capable of some learning and short- and long-term memory. The behaviors leading to successful metamorphosis are governed by a complex series of endocrine events.
In laboratory conditions, ants can combat a pathogen infection by means of the medicinal use of reactive oxygen species (ROS). However, it is still unknown where they obtain medicinal compounds in the wild and how they use them. Due to an upregulation of ROS in response to herbivory, aphid-infested plants have been suggested to be a potential source of ROS for ants in the wild. We investigated whether infection would cause Lasius platythorax ants to change their foraging on extrafloral nectar on aphid-infested plants. We found no clear evidence for the ants significantly changing their foraging behaviour in response to the pathogen, nor for the extrafloral nectar to contain ROS. The aphids in our experiment had a relatively high concentration of ROS and future research should determine whether predation on aphids could be a potential source of both protein and ROS needed to combat a disease.
Full-text available
Leaf-cutting ants depend on mutualisticfungi to survive. An infection that massively affects the workers compromising the proper maintenance of the fungus, or that can attack the fungus garden, can be fatal to the colony. Thus, leaf-cutting ants have evolved a complex defense system composed of both innate individual immunity and collective immunity to protect the colony against potential threats. To characterize the collective and individual immunity of Atta cephalotes workers to Metarizhium anisopliae we assessed the hygienic behavior and the expression of antimicrobial peptides of A. cephalotes workers triggered by Metarizhium anisopliae spores. As a control challenge, workers were treated with water. Regardless of whether the challenge was with water or spore suspension, A. cephalotes workers displayed an immediate response characterized by an increase in time spent both self-grooming and collective grooming along with a reduction in time spent fungus-grooming. The individual immunity triggered the expression of abaecin as early as 24 hours post-infection, exclusively in workers challenged with M. anisopliae. In contrast, the level of expression of defensin remained constant. These results suggest that upon being challenged with a suspension of M. anisopliae spores, A. cephalotes workers deploy both collective and individual immunity to produce a response against the invader. However, when the spores of M. anisopliae are applied as liquid suspension collective immunity deploys a generic strategy, while individual immunity shows a specific response against this entomopathogen.
The ecological success of ants relies on their high level of sociality and cooperation between genetically related nestmates. However, these group-living insects suffer from elevated risks of disease outbreak in the whole nest. To face this sanitary challenge, social and spatial distancing of pathogen-exposed individuals from susceptible nestmates appear to be simple, although efficient, ways to limit the propagation of contact-transmitted pathogens. Here we question whether spatial distancing in Myrmica rubra ants is an active response of diseased individuals that correlates with their level of infectiousness. We contaminated foragers with spores of Metarhizium brunneum entomopathogenic fungus. We daily tracked the location of these pathogen-exposed individuals and we analyzed their movement patterns until their death on the 5th day post-contamination. Quite unexpectedly, we found that contagious individuals, whose body was covered with infectious spores, did not reduce their mobility nor stayed far away from larvae in order to limit pathogen transmission to healthy nestmates. Spatial distancing occurred later when diseased individuals were no longer contagious because spores had penetrated their body. These sick ants mainly stayed outside the nest, were less mobile and showed a shift from a superdiffusive to subdiffusive walking pattern. Furthermore, these diseased ants did not actively head towards directions that were opposite to the nest entrance. This study found no evidence for early spatial distancing by contaminated M.rubra workers that would fit to the actual risk of colony-wide contagion. Coupled to a lower mobility and area-reduced walking patterns, the late distancing of moribund individuals appears to be a symptom of sickness resulting from fungus-induced physical and physiological dysfunctions. Besides questioning the truly altruistic nature of death in isolation in this system (and potentially others), we discuss about the ecological and physiological constraints that explain the absence of early distancing when some ant species are exposed to pathogens.
These proceedings contain 18 papers that discuss topics on speciation and adaptation; life history, evolution, phenotypic plasticity and genetics; sexual selection and reproductive biology; insect-plant interactions; insect-natural enemy interactions; and social insects. A series of empirical case studies in evolutionary ecology using insects as model systems are also presented.
A paradigm for the evolution of cooperation between parasites and their hosts argues that the mode of parasite transmission is critical to the long-term maintenance of cooperation. Cooperation is not expected to be maintained whenever the chief mode of transmission is horizontal: a parasite's progeny infect hosts unrelated to their parent's host. Cooperation is expected to be maintained if the chief mode of transmission is vertical: a parasite's progeny infect only the parent's host or descendants of that host. This paradigm was tested using bacteria and filamentous bacteriophage (f1). When cells harboring different variants of these phage were cultured so that no infectious spread was allowed, ensuring that all parasite transmission was vertical, selection favored the variants that were most benevolent to the host-those that least harmed host growth rate. By changing the culture conditions so that horizontal spread of the phage was allowed, the selective advantage of the benevolent forms was lost. These experiments thus support the theoretical arguments that mode of transmission is a major determinant in the evolution of cooperation between a parasite and its host.
Multiple mating by females (polyandry) remains hard to explain because, while it has substantial costs, clear benefits have remained elusive. The problem is acute in the social insects because polyandry is probably particularly costly for females and most material benefits of the behavior are unlikely to apply. It has been suggested that a fitness benefit may arise from the more genetically diverse worker force that a polyandrous queen will produce. One leading hypothesis is that the increased genetic diversity of workers will improve a colony's resistance to disease. We investigated this hypothesis using a polyandrous leaf-cutting ant and a virulent fungal parasite as our model system. At high doses of the parasite most patrilines within colonies were similarly susceptible, but a few showed greater resistance. At a low dose of the parasite there was more variation between patrilines in their resistance to the parasite. Such genetic variation is a key prerequisite for polyandry to result in increased disease resistance of colonies. The relatedness of two hosts did not appear to affect the transmission of the parasite between them, but this was most likely because the parasite tested was a virulent generalist that is adapted to transmit between distantly related hosts. The resistance to the parasite was compared between small groups of ants of either high or low genetic diversity. No difference was found at high doses of the parasite, but a significant improvement in resistance in high genetic diversity groups was found at a low dose of the parasite. That there is generic variation for disease resistance means that there is the potential for polyandry to produce more disease-resistant colonies. That this genetic variation can improve the resistance of groups even under the limited conditions tested suggests that polyandry may indeed produce colonies with improved resistance to disease.
This chapter provides general guidelines for the recognition, handling and initial diagnosis of diseased insects and the identification of major entomopathogen groups. Insects are associated with a broad diversity of microorganisms in a variety of symbiotic relationships including commensalism, mutualism, and parasitism. There is an astronomical number of entomopathogens which cause diseases and there is a great a number of insect hosts to find them. Both living and dead insects that are patently infected with entomopathogens can be found in virtually every setting inhabited by insects including natural terrestrial and aquatic ecosystems, agroecosystems and in laboratory and commercial insect colonies. The recognition of diseased insects in the field or subsequently in the lab will initially rely on gross pathology and patent infections. Insects that are patently infected with entomopathogens often manifest characteristic symptoms and signs of diseases. Color changes due to entomopathogens in living insects are usually associated with those insects with transparent to semitransparent integuments. It is suggested that when an insect is suspected of being infected with an entomopathogen, it should be examined as soon as possible after collection. The invasion of cadavers by fast-growing saprophytic organisms may complicate the determination of the true etiological agent. The general characteristics of insect disease caused by the major groups of entomopathogens are also elaborated.