Content uploaded by Volker Nehring
Author content
All content in this area was uploaded by Volker Nehring on May 17, 2019
Content may be subject to copyright.
Brood adoption in the leaf-cutting ant Acromyrmex
echinatior: adaptation or recognition noise?
B. Fouks1,2, P. d'Ettorre1,3 and V. Nehring1
1. Centre for Social Evolution, Department of Biology, University of
Copenhagen, Universitetsparken 15, 2100 Copenhagen E, Denmark
2. AG Molecular Ecology, Martin-Luther University of Halle-Wittenberg,
Hoher Weg 4, 06099 Halle, Germany
3. Laboratory of Experimental and Comparative Ethology (LEEC),
University of Paris 13, 99 av. J.B. Clément, 93430 Villetaneuse, France
Correspondence: Volker Nehring, Centre for Social Evolution, Department of
Biology, University of Copenhagen, Universitetsparken 15, 2100 Copenhagen E,
Denmark. Tel. +45 35 32 12 8; Fax +45 35 32 15 50; e-mail:
vnehring@bio.ku.dk
Word count: main text 3971, total 5560, 4 figures
1
2
3
4
5
6
7
10
11
12
13
14
15
16
17
18
19
20
21
22
23
1
Abstract
The ability to discriminate between nestmates and non-nestmates is an important
prerequisite for the evolution of eusociality. Indeed, social insect workers are
typically able to discriminate between nestmate and non-nestmate workers. Adult
non-nestmate workers are readily detected and rejected from the colony. Whether
social insects can discriminate between nestmate and non-nestmate brood,
however, is less clear. Here we show that workers of the leaf-cutting ant
Acromyrmex echinatior discriminate between nestmate and non-nestmate brood,
and among brood of different stages. Initially, non-nestmate brood is attacked, but
it is adopted after a delay. Adoption could occur due to inefficiency of the
recognition system, or it could be adaptive because it is an inexpensive way to
increase the workforce. Our results suggest that brood adoption may occur
accidentally. We also report how workers replace fungal hyphae on the brood's
surface before transporting the brood into their fungus garden.
Keywords: Acromyrmex; Brood care; Fungus-growing ants; Nestmate recognition
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
2
Introduction
Recognition is an important factor stabilizing all kinds of social interactions, from
reciprocity in repeated encounters, to preferential treatment of kin, or inbreeding
avoidance. In social insects, an elaborate nestmate recognition system protects the
colony resources from unrelated intruders (cf. d'Ettorre and Lenoir, 2010). Typical
intruders into ant colonies are social parasite queens, slave making ant workers, or
workers from other ant colonies. It is therefore crucial that ant workers can
discriminate between nestmate and non-nestmate workers. Social insect brood,
however, is generally helpless, not mobile, and not able to survive on its own even
for short periods. Thus allocolonial brood is presumably not a common intruder
into social insect colonies and far less a danger to an insect society than
allocolonial workers are. On the contrary, it has been shown that allocolonial
brood integrates into colonies without harming them, and the newly emerging
workers will work for the adopting colony (Carlin and Hölldobler, 1986; Hare and
Alloway, 1987; Carlin and Schwartz, 1989; Fénéron and Jaisson, 1995; Isingrini
et al., 1985). Adopting pupae, for example, would promote the growth of the
adopting colony, since it saves the resources the individual would have consumed
during its larval development. This effect is exploited to reinforce Oecophylla
colonies, which are used for biological control in fruit plantations in Asia (cf.
Krag et al., 2010). One might therefore be inclined to expect that workers will
adopt allocolonial brood, particularly at a late developmental stage. This pattern
has been described for instance in Lasius ants (Lenoir, 1981). On the other hand,
ants that are threatened by social parasites may benefit from efficiently
discriminating between nestmate and non-nestmate brood to hinder the
reproduction of parasites, which would be very costly to the host colony. The
question of whether or not there is discrimination against non-nestmate brood, and
how efficient it is, is therefore very relevant to understanding how colony integrity
and efficacy are traded off.
The evidence for colony-specific brood recognition in ants is conflicting. For
many species it is known that ants react differently to allo- and concolonial brood
(Meudec, 1978; Lenoir, 1984; Bonavita-Cougourdan et al., 1987; Fénéron and
Jaisson, 1992). Leaf-cutting ants have been reported to discriminate nestmate
from non-nestmate brood as well, but also to adopt a high proportion of non-
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
3
nestmate brood (Febvay et al., 1984; Viana, 1996; Viana et al., 2001; de Souza et
al., 2006). All studies have recorded only acceptance or rejection of brood, so that
it was not possible to examine the reasons for adoption in detail. Leafcutters grow
a fungus garden as a food resource, in which the brood lives. The ants, their
brood, and fungus have coevolved, leading to morphological and functional
adaptations by larvae to feed on gongylidia, special nutrient-rich hyphal tips. A
larva can only survive if workers place the gongylidia on the larva's mouthparts
and keep the larva clean (Wheeler, 1948; Weber, 1966). Furthermore, workers
plant fungal hyphae on the brood's cuticle; the purpose of this behaviour has not
yet been brought to light (Lopes et al., 2005). The brood is likely to communicate
its need for food or other care via semiochemicals, and these may interfere with
the chemical cues that are used for nestmate recognition (Morel and Vander Meer,
1988).
We studied the adoption of brood items by workers of the leaf-cutting ant
Acromyrmex echinatior. We hypothesized that adoptions could happen either
“intentionally”, because it is adaptive, or because of “noise”, i.e. errors in
nestmate recognition. It is possible that nestmate recognition of brood is not a
mechanism that has been selected for, since workers rarely, if ever, encounter
allocolonial brood. However, A. echinatior can be invaded by the inquiline social
parasite A. insinuator (Sumner et al., 2003) and in this case brood recognition
would be useful. In the case of same-species interactions, the adoption of later
brood stages would be more beneficial to a colony than that of early stages, since
the investment that has to be made before a worker ecloses is lower for older
brood stages. Therefore, if adoption of brood was intentional, one would expect a
higher propensity for workers to adopt pupae than larvae. We tested this by
presenting leafcutter workers in small subcolonies with brood of different stages.
We examined by detailed behavioural observations, whether the workers are able
to discriminate between nestmate and non-nestmate brood as well as different
brood stages.
Material and methods
Ten laboratory colonies of Acromyrmex echinatior were used for our experiments
(Ae168, Ae204, Ae331, Ae332, Ae334, Ae335, Ae345, Ae347, Ae356, and
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
4
Ae372). The colonies were collected from secondary rain-forest in Gamboa,
Panama, between 2002 and 2008. After being transferred to the laboratory in
Copenhagen, the colonies were maintained at constant temperature (25°C) and
humidity (75% relative humidity) on a diet of bramble leaves, apple, and rice.
To test for colony-specific brood recognition, con- and allocolonial brood items
were introduced into small discriminator subcolonies. These were housed in Petri
dishes (Ø 8 cm, 2 cm height) and contained a small fungus garden (12 cm³), 5
large workers, 5 medium workers, and 10 small workers (cf. Bot and Boomsma,
1996). A piece of bramble leaf (16 cm²) served as substrate for the fungus, from
which we had removed all brood.
From five of the ten original colonies, we created six discriminator subcolonies
each. These were divided into three pairs, each of them receiving brood of a
different stage: one pair received pupae (3.5 - 5.5 mm long; 1 - 2.5 mm wide),
another one large larvae (3.5 - 4.5 mm long; 1.5 - 2.5 mm wide), and the third pair
received medium sized larvae (2.5 - 3.5 mm long; 1 - 1.5 mm wide). The brood
that two discriminator subcolonies of a pair received stemmed from different
sources: one received brood from the same original colony the workers and
fungus came from (concolonial), while the other one received brood from one of
the five colonies that were not used to create discriminator subcolonies
(allocolonial).
Five brood items of the same stage and from the same colony were introduced
into each discriminator subcolony sequentially, at intervals of 30 minutes, starting
one hour after setting up the subcolonies. The brood items were introduced
through a hole in the lid of the Petri dish. Following the introduction, the
behaviour of the discriminator workers towards each brood item was recorded
during 15 minutes or until the ants brought the brood into the fungus garden. The
duration of antennation, grooming (i.e. licking and other manipulation of the
brood with mouth parts, excluding carrying), and aggression (i.e. rapid movement
towards the brood with mandible opening (threat), sometimes followed by biting)
were recorded with the software EthoLog (Ottoni, 2000). The observer was
always blind to the origin (concolonial / allocolonial) of brood items. In total we
observed the behaviour of five brood items per discriminator subcolony, with six
subcolonies (30 individual brood items) per original colony. As we used five
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
5
original colonies, we observed 150 brood items in total, namely 50 individual
brood items per each of the three brood stages.
The behavioural data were analysed with generalized linear mixed models (Bates
et al., 2008; R Development Core Team, 2010) including the colony origin as a
random factor. The stage of brood items (factor “stage” with three levels pupa,
large larva, medium larva) and the combination of brood and discriminator (factor
“origin” with two levels allocolonial and concolonial) were included as fixed
factors in all models. We tested with likelihood ratio tests, which of the factors
had a significant influence on the response variable, by subsequently removing
factors and interactions between factors from a full model that included both
factors and their interaction.
The distributions of all response variables and their residuals were inspected for
symmetry, and the variables were transformed when indicated. The durations of
antennation and grooming were log-transformed and then analysed using models
with Gaussian error families and identity link functions. As aggression was rare,
we transformed the variable into a binomial distribution (presence or absence of
aggression during the observation of a single brood item) and analysed it with the
respective error family and logit link function. To analyse whether a single
antennation is enough for workers to identify the stage and origin of brood, we
compared the distribution of the first behaviour following the initial antennation
with Pearson's Chi squared test.
We also recorded when a brood item was first contacted by a worker, and when it
was brought into the fungus garden. We calculated the delay t(d) from the first
contact t(fc) with the brood item until it arrived in the fungus garden t(t), as t(d) =
t(t) – t(fc). Brood items that were not brought into the fungus garden during the
first 15 minutes were assigned the value of the total observation time (t(t) = 900
sec). This variable was analysed assuming a Gaussian distribution. We also noted
whether brood was constantly contacted by workers after they discovered it and
before they brought it into the fungus garden, and analysed these observations as a
binomial variable. “Constant contact” thus means that the brood item received
continuous antennation and/or grooming from the first discovery to the transport
into the fungus garden.
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
6
The number of brood items that were transported into the nest during the 15
minutes of observation was analysed with Poisson error families and log link
function. One hour, 3h, and 24h after this, the number of brood items inside the
fungus garden was recorded and analysed in a similar way, but adding time as a
covariate.
The amount of fungal hyphae covering the brood items was recorded before the
start of the experiments and after 24 hours. We defined four levels of fungus
coverage: (0) no fungus hyphae covering pupa or larva, (1) one to six fungus
hyphae on pupa or larva, (2) more than 1/3 and less than 2/3 of the pupal or larval
surface covered by fungus, (3) more than 2/3 of pupa or larva covered by fungus.
The fungus coverage was analysed with Poisson errors and using time (before /
after the experiment) as a third explanatory factor.
Results
Overall, allocolonial brood received more antennation than concolonial brood (χ²1
= 7.1, p < 0.01), and medium larvae were antennated less than large larva and
pupae (χ²2 = 19.3, p < 0.001, Fig. 1a). There was no interaction between the
factors (χ²2 = 3.2, p = 0.2). Workers spent more time grooming allocolonial than
concolonial larvae (Fig. 1b). Allocolonial pupae, however, where groomed less
than concolonial pupae. Comparing among concolonial brood items, pupae were
groomed more than larvae; however, this pattern is not present among allocolonial
brood items. The interaction between stage and colony origin was significant
(origin x stage χ²2 = 12.8, p < 0.01; main effects: origin χ²1 = 0.8, p = 0.37; stage
χ²2 = 19.88, p < 0.001).
The distribution of behavioural acts succeeding initial contact by workers differed
between allocolonial and concolonial brood items (χ²3 = 12.5, p < 0.01). While
concolonial brood items were mostly groomed, this behaviour was rarer for
allocolonial brood, which was ignored or even attacked. There was no difference
in behaviour towards brood of different stages (χ²3 = 2.3, p = 0.5). We also
compared allo- and concolonial brood for each stage separately (pupae χ²3 = 7.4, p
= 0.06; large larvae χ²3 = 6.4, p = 0.10; medium larvae χ²3 = 6.0, p = 0.11), and
among the three stages for con- or allocolonial brood separately (concolonial χ²6 =
14.2, p < 0.05; allocolonial χ²6 = 10.1, p = 0.12), and did not find any clear trend
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
7
in the behaviour that followed the first contact of workers with brood items, other
that workers might be more likely to detect the difference between different brood
stages when the brood is concolonial rather then allocolonial.
During the 15 minutes of observation, concolonial brood items were brought into
the fungus garden quicker (t(d)) than allocolonial ones (Fig. 1c). Considering
allocolonial brood, medium sized larvae were brought into the fungus garden
much quicker than large larvae and pupae, while concolonial larvae of both size
classes were in the fungus garden earlier than concolonial pupae (interaction
origin x stage χ²2 = 6.2, p < 0.05; origin χ²1 = 37.8, p < 0.001; stage χ²2 = 39.4, p <
0.001).
Whether a brood item was in “constant contact” with the discriminator workers
after its first discovery depended on the brood's origin and stage and their
interaction (origin x stage χ²2 = 9.3, p < 0.01; origin χ²1 = 11.2, p < 0.001; stage χ²2
= 24.4, p < 0.001). While 60 % of the large concolonial larvae were continuously
tended, this was the case for only 8 % of the large allocolonial larvae (Fig. 2a).
Medium larvae received most attention overall and there was only a small
difference (80 % vs 60 %) between con- and allocolonial medium larvae. Only 28
% of both concolonial and allocolonial pupae were tended continuously by the
discriminator workers.
Only allocolonial brood items (12 % - 24 %) received aggressive acts by workers
(χ²1 = 19.4, p < 0.001, Fig. 2b). Neither the stage of the brood (χ²2 = 1.3, p = 0.53)
nor its interaction with the origin (χ²2 < 0.1, p > 0.99) explained any of the
variance.
After the first 15 minutes of continuous observation, the ants kept on transporting
brood into the fungus garden, so that eventually all five brood items were in the
garden in all experiments (Fig. 3). However, allocolonial brood was brought into
the garden later than concolonial brood (interaction time x origin χ²1 = 8.1, p <
0.01). The stage of a brood item did not influence when it was brought into the
fungus garden, neither on its own nor in interaction with other variables (origin x
stage x time χ²2 = 1.2, p = 0.55; stage x origin χ²2 = 2.8, p = 0.24; stage x time χ²2
= 2.15, p = 0.34; stage χ²2 = 3.6, p = 0.16; origin χ²1 = 82.9, p < 0.001; time χ²1 =
65.6, p < 0.001).
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
8
Before the experiment, pupae were covered by more fungal hyphae than larvae. At
the end of the experiment, 24 hours after the introduction of brood items into
discriminator subcolonies, both larvae and pupae were covered by more hyphae
than before the experiment (Fig. 4). It did not matter whether the brood was allo-
or concolonial, as the factor origin and its interactions with other factors did not
contribute to the model (origin x stage x time χ²2 = 0.4, p = 0.82; origin x stage χ²2
= 2.1, p = 0.35; origin x time χ²1 = 1.2, p = 0.26; origin χ²1 = 2.5, p = 0.11). The
variation in the brood's fungal coverage was determined by an interaction of stage
and time only (stage x time χ²2 = 28.2, p < 0.001; stage χ²2 = 180.2, p < 0.001;
time χ²1 = 59.9, p < 0.001), underlining that pupae gained relatively more fungal
coverage than larvae.
Discussion
Our results demonstrate that Acromyrmex echinatior workers are able to
discriminate not only nestmate from non-nestmate brood, but also brood of
different stages. Although non-nestmate brood was rejected at first, all brood
items were transported into the fungus garden within 24 hours. We present
evidence indicating that the adoption is taking place by accident rather than
because it is adaptive.
Colony-specific brood recognition in Acromyrmex echinatior is clearly
demonstrated by the fact that only allocolonial brood received aggression. When
allocolonial brood was not attacked, it was ignored after the first antennation;
moreover, allocolonial brood was transported to the fungus garden significantly
later than concolonial brood. Although allocolonial brood was initially aggressed
or ignored, all brood items have been brought into the nest within 24 hours. This
is surprising at the first glance, but among leaf-cutting ants it is not uncommon
(Febvay et al., 1984; Viana, 1996; de Souza et al., 2006). There are two different
explanations for this phenomenon.
First, it is possible that adopting allocolonial brood is not per se harmful to a
colony. If workers that grow up in a foreign colony will integrate well and do not
jeopardize the colony's fitness, then it is adaptive to adopt allocolonial larvae and
pupae, as it for instance occurs in species that perform intra-specific brood raids
(e.g. Rissing and Pollock, 1987; Kronauer et al., 2003). A colony needs to invest
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
9
less into these foreign workers than into their own, since they save the investment
already made by the foreign colony before adoption. Indeed, there is sufficient
evidence suggesting that non-nestmate workers integrate well into foreign
colonies and adoptions can be successful (Isingrini et al., 1985; Hare and Alloway,
1987; Carlin and Schwartz, 1989; Fénéron and Jaisson, 1995;). An alternative
hypothesis would be that workers adopt allocolonial brood by mistake, because
they cannot discriminate it from concolonial brood. As allocolonial brood does
typically not harm the colony, there may generally be only little selection pressure
on ants to detect non-nestmate brood. However, when the colony is at risk to be
infected by social parasite ants, it will be adaptive to be able to identify the social
parasite brood. Acromyrmex echinatior is host to the social parasite Acromyrmex
insinuator, and the host workers seem to be able to discriminate parasite from host
brood under laboratory conditions (Sumner et al., 2003).
Two of our observations support the “mistaken adoption” hypothesis and
contradict the “adaptive adoption” hypothesis. First, under the “adaptive
adoption” hypothesis, one would expect that workers adopted pupae and large
larvae more readily than medium larvae, because the latter require a higher
investment before they can be useful to the colony. We found the opposite:
medium size larvae are accepted more readily than larger and likely older larvae
and pupae, which is consistent with results in Acromyrmex subterraneus (Viana,
1996). Large larvae had the same low levels of fungus coverage as medium
larvae, and therefore the observed delayed adoption of large larvae cannot be due
to workers spending more time to remove fungal hyphae before transporting the
larvae into the nest.
Second, allocolonial brood was brought to the nest later than concolonial brood.
This reflects that allocolonial brood was only adopted after a delay, the reason for
this might be that the nestmate recognition label of the brood changed over time.
Social insects typically discriminate against non-nestmates using a colony-specific
label, encoded in the chemical profile located on the cuticle (Lenoir et al., 1999;
d'Ettorre and Lenoir, 2010). The nestmate recognition labels are dynamic: during
allogrooming and trophallaxis, chemicals are transferred between ants. In this way
inter-individual differences in labels are aligned, which guarantees that all
nestmates share the same nestmate recognition label (Lenoir et al., 1999; Hefetz,
2007). During the experiment, the brood was always close to the fungus garden
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
10
and came into contact with the discriminating workers repeatedly. This could have
allowed transfer of cuticular chemical substances (Soroker et al., 1995). Thus,
even though workers correctly identified allocolonial brood as non-nestmate in the
beginning of the experiment, the chance of mistaking allocolonial brood for
nestmate may have increased with time because the label of allocolonial brood
was aligned to that of the discriminator colony. It has been shown that the
cuticular chemical profiles of brood of leaf-cutting ants, including A. echinatior,
are rather simple as compared to that of adult workers (Viana et al., 2001; Richard
et al., 2007), in that they consist of less substances and mainly of saturated
alkanes, which are generally assumed to play a minor role in nestmate recognition
(review in d’Ettorre and Lenoir 2010; van Zweden and d’Ettorre, 2010). It is
therefore possible that the brood is chemically transparent (cf. Martin et al., 2008),
i.e. it does not bear any nest-specific odour and so there is no mechanism of
colony-specific brood recognition. Interactions between brood and workers would
then be based on other semiochemicals, for example brood pheromones (Morel
and Vander Meer, 1998) and/or specific behaviours, such as begging. We can
exclude a total chemical transparency of A. echinatior brood because initially non-
nestmate brood is detected; the colony-specific profile may however be very weak
and thus fade away or be replaced fairly quickly during our experiments, which
would assist the unintended adoption process.
The alignment of cuticular profiles could also have been facilitated by workers
replacing the brood's fungal cover, as the fungus possesses a colony-specific
chemical profile that allows workers to discriminate against non-nestmate fungus
(Viana et al., 2001; Richard et al., 2007). Little is known about the exchange of
odours between workers or fungus and brood, but it seems that there is a relation
between fungus and brood chemical profiles since the two are similar to each
other (Viana et al., 2001; Richard et al., 2007). It has been noted before that
leafcutter brood is often covered by fungal hyphae, which nurse workers plant on
the surface of larvae and pupae (Lopes et al., 2005). It is not clear what the benefit
of fungal hyphae growing on brood is for the ants. In undisturbed colonies we
recorded that pupae were covered by hyphae to a greater extent than larvae, which
often do not have fungal cover. At the end of our experiments, the fungal cover
particularly of larvae, but also pupae, had increased. It might well be that brood
outside the fungus garden is likely to be contaminated by pathogens; thus workers
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
11
clean the brood thoroughly and plant new hyphae before they transport it into
their nest, to avoid a contamination of their fungus garden.
What we recorded as “grooming” was in fact a composite behaviour. Workers did
not only lick brood, which protects larvae against diseases (Hughes et al., 2002);
we also observed that workers removed the fungus cover and planted new hyphae
before they carried a brood item into the fungus garden. The fact that more fungal
hyphae had to be removed from pupae than from larvae might explain why larvae
were generally brought into the fungus garden earlier than pupae. It is interesting
that it did not matter whether the brood was allo- or concolonial; the hyphal cover
of all brood items was replaced, irrespective of its origin.
We can only assume that replacing the hyphal cover serves to avoid contamination
that might threaten the fungus garden, which is a resource of vital importance for
the ant colony (Currie, 2001; Poulsen et al., 2005). A thorough cleaning of brood
is only possible when the brood's surface is free from hyphae; planting hyphae on
all brood items after cleaning might help to maintain the brood hygienic. It is
possible that the hyphal cover may be a permanent protection against pathogens;
the fungus may not only be a physical barrier preventing pathogens from attacking
the brood, but it may also secrete antimicrobial substances (cf. Lopes et al., 2005).
In summary, we show that Acromyrmex workers discriminate between different
brood stages as well as between allo- and concolonial brood of all stages.
Allocolonial brood is adopted after a delay, which indicates that allocolonial
brood is adopted by mistake, possibly because the profile of brood is aligned to
that of the workers. This alignment may be supported by the removal and
subsequent planting of fungal hyphae on the brood's cuticle, which is carried out
for all brood items before they are brought to the fungus garden.
Acknowledgements
This work was supported by the Marie Curie Excellence Grant CODICES (EXT-
CT-2004-014202) and partly by a ‘‘Freia grant’’ from the Faculty of Science,
University of Copenhagen, both assigned to PdE. BF was supported by the
ERASMUS programme, and VN by the German Academic Exchange Service
(DAAD). We thank the Smithsonian Tropical Research Institute (STRI) and the
Autoridad Nacional del Ambiente (ANAM) for permission to work in Panama,
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
12
and the Danish Natural Science Research Council for financial support. The
authors also thank Hermogenes Fernandez Marin for fruitful discussions and all
the members of the Centre for Social Evolution, University of Copenhagen, for a
stimulating working environment.
References
Bates D., Maechler M. and Dai B. 2008. Lme4: Linear mixed-effects models
using S4 classes. R package version 0.999375-28. http://lme4.r-forge.r-
project.org/
Bonavita-Cougourdan A., Clément J.L. and Lange C. 1987. Nestmate recognition:
The role of cuticular hydrocarbons in the ant Camponotus vagus Scop. J.
Entomol. Sci. 22: 1-10
Bot A.N.M. and Boomsma J.J. 1996 Variable metapleural gland size-allometries
in Acromyrmex leafcutter ants (Hymenoptera: Formicidae). J. Kansas Entomol.
Soc. 69: 375-383
Carlin N.F. and Hölldobler B. 1986. The kin recognition system of carpenter ants
(Camponotus spp.): I. hierarchical cues in small colonies. Behav. Ecol. Sociobiol.
19: 123-134
Carlin N.F. and Schwartz P.H. 1989. Pre-imaginal experience and nestmate brood
recognition in the carpenter ant, Camponotus floridanus. Anim. Behav. 38: 89-95
Currie C.R. 2001. A community of ants, fungi, and bacteria: a multilateral
approach to studying symbiosis. Annu. Rev. Microbiol. 55: 357-380
d'Ettorre P. and Lenoir A. 2010. Nestmate recognition. In: Ant Ecology (Lach L.,
Parr C. and Abbott K., Eds),. Oxford University Press, USA. pp 194-208
de Souza D.J., Della Lucia T.M.C., Errard C., Richard F. and Lima E.R. 2006.
Behavioural and chemical studies of discrimination processes in the leaf-cutting
ant Acromyrmex laticeps nigrosetorus (Forel, 1908). Braz. J. Biol. 66 : 863-871
Febvay G., Mallet F. and Kermarrec A. 1984. Attractivité du couvain et
comportement des ouvrières de la fourmi attine Acromyrmex octospinosus (Reich)
(Hym. Form.). Actes Coll. Insectes Soc. 1: 79-86
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
13
Fénéron R. and Jaisson P. 1992. Nestmate-brood recognition among workers of
different social status in Ectatomma tuberculatum Olivier (Formicidae,
Ponerinae). Behav. Proc. 27: 54-52
Fénéron R. and Jaisson P. 1995. Ontogeny of nestmate brood recognition in a
primitive ant, Ectatomma tubercalutum Olivier (Ponerinae). Anim. Behav. 50: 9-
14
Hare J.F. and Alloway T.M. 1987. Early learning and brood discrimination in
leptothoracine ants (Hymenoptera: Formicidae). Anim. Behav. 35: 1720-1724
Hefetz A. 2007. The evolution of hydrocarbon pheromone parsimony in ants
(Hymenoptera: Formicidae) - interplay of colony odor uniformity and odor
idiosyncrasy. A review. Myrmecol. News 10: 59-68
Hughes W.O.H., Eilenberg J. and Boomsma J.J. 2002. Trade-offs in group living:
transmission and disease resistance in leaf-cutting ants. Proc. R. Soc. B 269: 1811-
1819
Isingrini M., Lenoir A. and Jaisson P. 1985. Preimaginal learning as a basis of
colony-brood recognition in the ant Cataglyphis cursor. Proc. Natl Acad. Sci.
USA 82: 8545-8547
Krag K., Lundegaard R., Offenberg J., Nielsen M.G. and Wiwatwittaya D. 2010.
Intercolony transplantation of Oecophylla smaragdina (Hymenoptera:
Formicidae) larvae. J. Asia Pacific Entomol. 13: 97-100
Kronauer D.J.C., Gadau J. and Hölldobler B. 2003 Genetic evidence for intra- and
interspecific slavery in honey ants (genus Myrmecocystus). Proc. R. Soc. Lond. B
270: 805-810
Lenoir A. 1981. Brood retrieving in the ant, Lasius niger L. Sociobiology 6: 153-
178
Lenoir A. 1984. Brood-colony recognition in Cataglyphis cursor worker ants
(Hymenoptera: Formicidae). Anim. Behav. 32: 942-944
Lenoir A., Fresneau D., Errard C. and Hefetz A. 1999. Individuality and colonial
identity in ants. In: Information Processing in Social Insects (Detrain C.,
Deneubourg J.L. and Pasteels J., Eds),. Birkhäuser Verlag, Basel. pp 219-237
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
14
Lopes J.F.S., Hughes W.O.H., Camargo R.S. and Forti L.C. 2005. Larval isolation
and brood care in Acromyrmex leaf-cutting ants. Insect. Soc. 52: 333-338
Martin, S.J., Takahashi, J.-I., Ono, M., and Drijfhout, F.P. 2008. Is the social
parasite Vespa dybowskii using chemical transparency to get her eggs accepted? J.
Insect Physiol. 15: 700-707
Meudec M. 1978. Response to and transport of brood by workers of Tapinoma
erraticum (Formicidae; Dolichoderinae) during nest disturbance. Behav. Proc. 3:
199-209
Morel L. and Vander Meer R.K. 1988 Do ant brood pheromones exist? Ann.
Entomol. Soc. Am. 81: 705-710
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
Poulsen M. and Boomsma J.J. 2005. Mutualistic fungi control crop diversity in
fungus-growing ants. Science 307: 741-744
R Development Core Team 2010. R: A Language and Environment for Statistical
Computing. R Foundation for Statistical Computing, Vienna, Austria. 1706 pp
Richard F., Poulsen M., Hefetz A., Errard C., Nash D.R. and Boomsma J.J. 2007.
The origin of the chemical profiles of fungal symbionts and their significance for
nestmate recognition in Acromyrmex leaf-cutting ants. Behav. Evol. Sociobiol. 61:
1637-1649
Rissing S.W. and Pollock G.B. 1987 Queen aggression, pleometrotic advantage
and brood raiding in the ant Veromessor pergandei (Hymenoptera: Formicidae).
Anim. Behav. 35: 975-981
Schönrogge, K., Wardlaw, J.C., Peters, A.J., Everett, S., Thomas, J.A. and Elmes,
G.W. 2004. Changes in chemical signature and host specificity from larval
retrieval to full social integration in the myrmecophilous butterfly Maculinea
rebelli. J. Chem. Ecol. 30: 91-107
Soroker V., Vienne C. and Hefetz A. 1995. Hydrocarbon dynamics within and
between nestmates in Cataglyphis niger (Hymenoptera: Formicidae). J. Chem.
Ecol. 21: 365-378
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
15
Sumner S., Nash D.R. and Boomsma J.J. 2003 The adaptive significance of
inquiline parasite workers. Proc. R. Soc. B 270: 1315-1322
van Zweden, J.S. and d'Ettorre, P. 2009. The role of hydrocarbons in nestmate
recognition. In: Insect Hydrocarbons: Biology, Biochemistry and Chemical
Ecology (Blomquist, G.C. and Bagnères, A.-G., Eds.), Cambridge University
Press, Cambridge, USA. pp 222-243
Viana A.M.M. 1996. La reconnaissance coloniale du couvain et du champignon
chez la fourmi champignonniste Acromyrmex subterraneus subterraneus. Thesis
University of Paris XIII.
Viana A.M.M., Frézard A., Malosse C., Della Lucia T.M.C., Errard C. and Lenoir
A. 2001. Colonial recognition of fungus in the fungus-growing ant Acromyrmex
subterraneus subterraneus (Hymenoptera: Formicidae). Chemoecology. 11: 29-36
Weber N.A. 1966. Fungus-growing ants. Science 153: 587-604
Wheeler G.C. 1948. The larvae of the fungus growing ants. Am. Mid. Natr. 40:
664-689
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
16
Figures
Fig. 1: The total duration of antennation (a) and grooming (b) that brood items of
different stages (ML: medium large; LL: large larvae; P: pupae) received by
concolonial (white boxes) and allocolonial (grey boxes) workers, and the delay
from first contact by the workers to the transport into the nest (c). Boxes cover the
interquartile range and whiskers all data points; the median is indicated by a
horizontal bar. Each box represents measurements from a total of 25 brood items
from 5 different colonies.
470
471
472
473
474
475
476
477
17
Fig. 2: Number of brood items that received constant attention (a) and that were
attacked (b) by discriminator workers. In total 25 concolonial (white) and 25
allocolonial (grey) medium larvae (ML) and the same number of large larvae (LL)
and pupae (P) from five different colonies were observed.
478
479
480
481
18
Fig. 3: Number of brood items inside the fungus garden over time. Concolonial
brood is brought into the fungus garden quicker than allocolonial brood, but after
one day all brood items are in the fungus garden. Because there were no
differences between the brood stages, items of different stages are pooled,
resulting in 15 con- and 15 allocolonial replicates per point in time. Each replicate
included five brood items. Continuous line: concolonial brood items, dashed line:
allocolonial brood items. Mean (circles) and standard errors are shown.
482
483
484
485
486
487
488
19
Fig. 4: Level of fungus coverage of medium larvae (circles, solid line), large
larvae (squares, dashed line), and pupae (diamonds, dotted line) before and after
the experiment. A high score indicates that most of the brood's cuticle is covered
by fungal hyphae. Since there was no difference in the fungal coverage of con-
and allocolonial brood, these groups are pooled, resulting in a sample size of 50
brood items from 5 different colonies per data point. The data points indicate the
mean values and the vertical lines the standard error.
489
490
491
492
493
494
495
20