1989; Hölldobler and Wilson, 1990). Members of the genusEpimyrma
are parasites of variousLeptothoraxspecies. FourEpimyrmaspecies,
E. algeriana, E. bernardi,E. ravouxiand E. stumperi, conduct well-
organized slave raids that involve group recruitment, using processions
and effective sting fighting.Epimyrma krausseiproduces few workers
and their raids are rare and ineffectual, whereasEpimyrma corsicaand
Epimyrma adlerziare workerless inquilines. A corresponding progression
from mating flights to within-nest mating is also observed in these
species. Colony foundation byEpimyrma queens follows the typical
pattern seen in many temporary social parasites and obligatory slave
makers, and involves the forceful usurpation of a host colony. However,
the method of invasion varies greatly among species.E. krausseiqueens
approach host colonies in an aggressive manner. Once the parasite
penetrates the host nest, she kills the host queen and is adopted by the
host workers. The queens ofE. ravouxiuse a more ‘conciliatory’ approach
initially, grooming and stroking host workers, but, once inside the nest,
the parasite mounts the host queen from behind and kills her by seizing
her around the neck with her sabre-shaped mandibles. E. stumperi
queens crouch down, freeze and seem to feign death during their initial
encounters with host workers, but subsequently begin to mount workers
from behind and to groom them, perhaps acquiring chemical recognition
cues in the process. However, once inside the nest, the parasite queen
systematically eliminates the host queens by mounting them, rolling them
over and grasping their necks in her mandibles until they succumb.
A similar evolutionary progression from dulosis to inquilinism might
also have occurred in the myrmicine genusStrongylognathus. However,
inStrongylognathus testaceus, the only known inquiline, the host queen
is apparently tolerated (see Hölldobler and Wilson, 1990, and references
therein).Conclusions
Studies of the behavioural ecology of social parasitism among ant species
have revealed an array of complex relationships, led to the discovery of
diverse mechanisms by which these relationships are initiated and main-
tained and provided insight into how these relationships initially evolve
and become further elaborated. Continuing research is likely to reveal
many additional socially parasitic species, new variations in life histories
and novel behavioural and chemical strategies of social exploitation.
Moreover, our understanding of the evolution of social parasitism in ants
is far from complete, and much remains to be determined regarding the
broader ecological causes, constraints and consequences of this phen-
omenon. Some of the current issues, debates and outstanding questions
regarding the evolution of social parasitism in ants include the following.
What pre-existing species characteristics facilitate, promote or are
necessary for the evolution of various forms of social parasitism? Under332 R.J. Stuart