and colony reproduction by ‘budding’, could also be a factor, since it
might facilitate the evolution of parasitic colony foundation by ensuring
large numbers of easily invaded nests. Moreover, as mentioned above, the
ability to learn relatively diverse nest-mate recognition cues might also be
necessary to facilitate the integration of alien individuals into colonies.
Taken together, these factors might explain why social parasitism appears
to be predominantly a temperate-zone phenomenon and why certain
taxonomic groups among the ants with certain arrays of traits might be
much more prone to the formation of socially parasitic relationships
among closely related species than are other groups.
The various considerations discussed above could facilitate cross-
species adoption and the evolution of the specialized forms of parasitic
colony foundation that characterize temporary social parasites, inquilines
and obligatory slave makers. The evolution of the various behavioural and
chemical tactics used in colony usurpation, as described above, are some-
what enigmatic but are probably based on the various mechanisms that
exist within species in the context of queen dominance, reproductive
competition and queen-number regulation in polygynous species
(Heinze, 1990, 1992; Buschinger and Heinze, 1992). However, a complete
explanation for the evolution of obligatory slavery in ants also requires an
explanation for the evolution of slave raiding, a behaviour that appears to
be exceedingly complex and highly specialized. In this respect, three
hypotheses have been advanced: the predation hypothesis, the territorial
hypothesis and the transport hypothesis (Buschinger, 1986; Hölldobler
and Wilson, 1990).
Darwin (1859) first advanced the predation hypothesis and suggested
that the ancestral species to slave makers might have been predatory
species that attacked other colonies to obtain food. In this case, the first
slaves would be prey items that were not consumed, managed to eclose in
the predator’s nest and were accepted into the colony. The territorial
hypothesis suggests that aggressive territorial interactions among colonies
of the same or closely related species were the evolutionary precursors to
slave raiding (Wilson, 1975b; Alloway, 1980; Stuart and Alloway, 1982,
1983). Territorial interactions often lead to opportunistic brood pilfering
and predation; and, again, captured brood that is not consumed might
survive to be the first slaves. The transport hypothesis suggests that slave
raiding evolved from the regular transport of brood among the multiple
nests of a single colony (i.e. a polydomous colony). According to this
scenario, attempts to transport brood between more distant and increas-
ingly less familiar nests of a single polydomous colony might give rise
to aggression and ultimately lead to the evolution of slave raiding
(Buschinger, 1970, 1986).
In comparing the three hypotheses presented above, the territorial
hypothesis appears to provide the most complete and reasonable explan-
ation for how slave raiding in ants might have evolved from the known
behaviour patterns of free-living ancestral species. Indeed, territorial
behaviour, as observed in free-living species encompasses important330 R.J. Stuart