Integration raids in the Amazon ant Polyergus rufescens (Hymenoptera, Formicidae)
ABSTRACT Groups of enslaved Formica fusca workers from mixed colonies of Polyergus rufescens with numerous slave workforce tend to split off and found small and almost homospecific nests around the main nest, with at least some of them connected with the latter with underground passages. Their inhabitants are able, at least temporarily, to adopt young F. fusca gynes. P. rufescens invades these satellite nests in a manner similar to the normal slave raids, and carries the slaves back to the main nest. The supposed evolutionary cause of this behaviour is to keep integrity of mixed colonies and prevent possible emancipation of slaves.
Summary. Groups of enslaved Formica fusca workers from
mixed colonies of Polyergus rufescens with numerous slave
workforce tend to split off and found small and almost homo-
specific nests around the main nest, with at least some of
them connected with the latter with underground passages.
Their inhabitants are able, at least temporarily, to adopt
young F . fuscagynes. P . rufescensinvades these satellite nests
in a manner similar to the normal slave raids, and carries the
slaves back to the main nest. The supposed evolutionary
cause of this behaviour is to keep integrity of mixed colonies
and prevent possible emancipation of slaves.
Key words: Social parasitism, mixed colonies, colony organ-
isation, slave emancipation, Formica fusca.
The most frequent slave species of P . rufescens in Central
Europe is Formica fusca, principally a mono- or slightly
polygynous and monocalic species. Its workforce usually
numbers from several hundred to a thousand, rarely reaching
1,500–2,000 adults in particularly big colonies (Savolainen,
1990; Seifert, 1996). Swarms of P . rufescens may number up
to about 2,000 workers (see below), and the proportion of
slaves in its mixed colonies reaches 80–90% (Seifert, 1996).
This means that a P . rufescens nest may support between
8,000 and 18,000 enslaved workers, what exceeds several
times the common size of a free-living colony of F . fusca.
Thus, the multiplication of swarm size typical of F . fusca
caused by P . rufescens is a peculiar experiment of nature. The
response of F . fusca to this situation has revealed a peculiar
element of P . rufescens biology.
The phenomenon was noticed and investigated during
observations of two big colonies of P . rufescens mixed with
F . fusca in the Białowieża Forest (NE Poland) in 2002 and
2003 in the periods of their raiding activity (late July/early
August). The raiding columns were estimated to comprise up
to 2,000 workers. Both colonies displayed the behaviour dis-
cussed, and it was especially spectacular in one of them.
The main nest of the mixed colony, inhabited by P . rufescens
and the majority of the slave swarm, was situated in a low
bank, with a few close entrances on its top. At the foot of
the bank, several satellite nests were scattered around an area
of nearly 5 m2. The largest distance between a pair of these
satellite nests was 3.2 m, and the distances between indi-
vidual satellite nests and the holes of the main nest ranged
from 1.6 to 4.4 m. At least some of these filial nests were con-
nected with the main one by underground tunnels. The entire
arrangement resembled a compound nest system of F . cinerea
(see e.g. Dlussky, 1967). Satellite nests were small and super-
ficial (in litter, under a stone or a piece of wood), each in-
habited by not more than a few dozen F . fusca workers. Some
of them included small groups of young sexual adults and
sexual pupae of P . rufescens. Sporadically, single P . rufescens
workers were seen in the satellite nests, but neither P .
rufescens worker pupae nor F . fusca enslaved pupae could be
The colony exhibited high raiding activity. It often and
successfully raided neighbouring F . fusca within a 77-metre
radius. Apart from usual distant raids, P . rufescens performed
short-distance raids on their own satellite nests. These pecu-
liar raids took place almost every day during the observation
periods (3 weeks each year), even when the weather was not
good enough to undertake a long-distance raid. Their organ-
isation was identical to that of typical raids – from the
recruitment phase to the outbound column reaching the tar-
get nest (see e.g. Mori et al., 1991). The difference appeared
when the “attackers” met the F . fusca workers, which were, in
fact, their colony mates. The ants would establish contact in
a friendly manner, palpating each other with antennae, which
was mainly initiated by F . fusca, noticeably astonished (but
not frightened) at that sudden inroad. The excitement of
Insect. Soc. 52 (2005) 103–104
© Birkhäuser Verlag, Basel, 2005
Integration raids in the Amazon ant Polyergus rufescens
Museum and Institute of Zoology, Polish Academy of Sciences, Wilcza 64, 00-679 Warsaw, Poland, e-mail: firstname.lastname@example.org
Received 18 August 2004; revised 27 September 2004; accepted 11 October 2004.
opposing hypotheses concerning the discriminative abilities
of P . rufescens come to mind. The first holds that there are
some slight differences between the odour signatures of F .
fusca workers from the main nest and those from the satellite
nests, and that P . rufescens is able to detect these nuances.
The first of these assumptions, however, seems to be hardly
probable, because of constant intense circulation of F . fusca,
caused by P . rufescens itself. The second hypothesis is that
P . rufescens does not distinguish its own, enslaved indivi-
duals from foreign, “free” ones of the same species, and that
it is ready to attack each nest of a dulotic species within its
range when the automated sequence of raid phases is com-
pleted, including a march (even the shortest) of the organised
column. If the target is their own satellite nest, the attackers,
fulfilling their robbery instinct, catch and take F . fusca
workers out, since there are no pupae and adults passively
submit themselves to it, with no resistance. This fits well the
old conjecture of Dobrzański and Dobrzańska (1978) that the
dulotic instinct of P . rufescens is directed not only at pupae
but (potentially) also at mature workers of slave species.
The biological sense of the described P . rufescens behav-
iour may lie in preventing disintegration of mixed colonies
by split-offs of slave groups and their possible emancipa-
tion due to conspecific queen adoption. Some observations
(Czechowski, unpubl.) suggest that such emancipation ac-
tually happens under certain environmental conditions. With
P . rufescens not counteracting this process, nests of the
former slaves might eventually surround its colony in a tight
These ‘integration’raids of European P . rufescens are the
second kind of intracolony raids known in this genus. Earlier,
so called ‘emigration’ raids were described in North Ameri-
can P . lucidus by Kwait and Topoff (1983). It may be
assumed, that the general mechanisms of both phenomena
are the same.
Dlussky, G.M., 1967. Murav’i roda Formika. Izd. Nauka, Moskva,
Dobrzański, J. and J. Dobrzańska, 1978. Some questions related to
mechanisms of slave-raids in amazon-ant Polyergus rufescens Latr.
Acta Neurobiol. Exp. 38: 353–359.
Kwait, E.C. and H. Topoff, 1983. Emigration raids by slave-making
ants: a rapid-transit system for colony relocation (Hymenoptera:
Formicidae). Psyche 90: 307–312.
Mori, A., D.A. Grasso and F. Le Moli, 1991. Eco-ethological study on
raiding behaviour of the European amazon ant, Polyergus rufescens
Latr. (Hymenoptera: Formicidae). Ethology 88: 46–62.
Savolainen, R., 1990. Colony success of the submissive ant Formica
fusca within territories of the dominant Formica polyctena. Ecol.
Ent. 15: 79–85.
Seifert, B., 1996. Ameisen: beobachten, bestimmen. Naturbuch
Verlag, Augsburg, 352 pp.
104 W. CzechowskiIntegration raids in Polyergus rufescens
P . rufescensworkers would decline from that moment and the
majority of them would start to withdraw, with only a small
group entering the “raided” nest. After a few minutes, these
individuals would go back towards the surface, carrying F .
fusca adults in their mandibles, and join the end of the
inbound column. In about 90% of cases, a carried slave was
seized by the mandible or by the antenna base, and it took on
the typical posture for adult transport in Formicaspecies. The
rest were taken hold of in every possible way (by the pedicel,
leg or antenna) and dragged along on the ground, never
attempting to free themselves.
There would be several such peculiar raids in one day,
their subsequent waning waves gradually merging with one
another, so that initially well organised raids would transform
into permanent individual activity of P . rufescens workers.
Usually, in the meantime, the main forces would make an
ordinary raid to a foreign F . fusca colony. Adult transport of
slaves from the satellite nests until dusk, i.e. 3–4 hours. Its
average efficiency was about a dozen of individuals/min.
(max about 30), which corresponds to 2–3 thousand slaves
carried from one or a few (very small) nests of the main nest
daily. The only reliable explanation for such intensity of
transport is that the transferred individuals came back imme-
diately or that they were being systematically replaced by an
influx of other workers to satellite nests from the main nest
through underground passages (there was no increased F . fus-
ca on the surface). It is also thinkable that the “attackers”
used these passages to reach their own nest and obtained their
“victims” from there. If so, however, one could expect that
they would walk off with enslaved pupae rather than with
During the observations, nuptial flights from neighbour-
ing free-living F . fusca colonies were going on, and young
gynes were seen on the ground now and then. Twice in 2002,
F . fusca gynes, apparently – at least temporarily – adopted by
enslaved F . fuscaworkers, were noticed in satellite nests of the
mixed colony. On 1stAugust, one such nest was dug up, and a
F . fusca microgyne was found inside. Unfortunately, it was
fatally wounded during the inspection. This nest was 3.2 m
away from the entrances of the main nest and it was connect-
ed with the latter by an underground passage. Then, on 3rd
August, a F . fusca macrogyne was seen for a while under a
stone covering another satellite nest, 3 m away from the main
nest entrances. It was accompanied by a dozen or so of F . fus-
ca workers; immediately after the nest was uncovered, the
gyne hid inside the ground. Its further fate is unknown.
The questions arise about the mechanisms underlying the
behaviour reported (proximate factors) and its biological
sense (ultimate factors). Regarding the mechanisms, two