282 
Psyche 
[Vol. 88 
Within the foraging radius of an established colony, there is likely to 
be less food available, even if the established colony ignores 
incipient colonies. The amount of depletion will depend on the 
amount of resource overlap. Because destruction of females and 
incipient colonies is frequently reported and resource depletion 
probably also affects colony persistence, the chance of a small 
colony becoming established is low. Wilson (1971) estimates that 
only 0.01% of all fertilized females survive to found successful nests. 
Therefore, established colonies tend to persist and interact over long 
periods, insofar as is known (Wilson 1971). Given this pattern, what 
is the form of the interaction and why are patterns of interspecific 
overdispersion so common? 
According to current theory, species can segregate a habitat to 
avoid or lessen competition in several ways: microhabit partition- 
ing, food size or type, and activity period (Pianka 1978). Further, 
equilibrium theory generally asserts that only a limited amount of 
overlap is tolerated on any given niche axis (Mac Arthur and Levins 
1967; Colwell and Futuyma 1971). Species which are too similar 
should not be able to coexist and, over a long enough period, the 
superior competitor in the overlapping pair will drive the other 
species extinct. Although there are many problems with the assump- 
tions of this argument, we will use its basic divisions to examine the 
patterns of overlap between co-occurring ant species. Ant species 
may be specialized along these three major axes. We will consider 
each potential kind of specialization in turn and evaluate the 
evidence that segregation of species along that factor is usually 
sufficient to prevent strong competitive interactions. 
Species may be considered specialized on food types in 3 major 
ways: (1) restricted prey types (i.e., only centipedes), (2) specific size 
ranges of prey (i.e., only prey 1-3 mm in length) or (3) some 
combination of (1) and (2) (i.e., centipedes between 5 and 8 mm 
long). Different kinds of specializations will have different effects on 
colony structure, nest size and foraging strategy. Resource restric- 
tion is frequently based on the matching of mandible or head size to 
food size or type (the trophic appendage, Schoener 1971, see below). 
Resources which are retrieved by individual workers, not by 
coordinated action, are especially likely to be treated in this manner 
(for example, seeds for desert ants, collembolans for dacetines). The 
resistance of the food item to recovery is also important; items 
