1981] 
Levings & Traniello — Territoriality in Ants 
287 
noda and Camponotus sp. in Kenyan forest; alarm/ recruitment 
specification may be the behavioral mechanism responsible for the 
structure and maintenance of the tropical canopy ant mosaic (Holl- 
dobler 1979b). 
In general, to defend any resource or area, including the nest, an 
ant must be able to summon her nestmates to a particular location. 
Within the nest, even quite primitive ants are able to communicate 
alarm and attract reinforcements (Robertson 1971; Traniello un- 
published data). Outside the nest, recruitment is a necessary com- 
ponent of effective defense. 
Ant species possess a wide diversity of recruitment communica- 
tion techniques that are ecologically significant (see review by 
Holldobler 1977). It is important to understand the ethology of 
social design to comprehend its role in ecological interaction. There 
are definite phylogenetic constraints and/or trends in the form of 
recruitment communication within the various subfamilies of ants 
(Holldobler 1977). More primitive groups (some Ponerinae) usually 
recruit few workers to food sources; some group raiding species are 
exceptions. Mass recruitment is characteristic of some groups of 
Myrmicinae, Dolichoderinae and Formicinae. Each lineage has 
developed within certain paths involving specific glandular, physical 
and behavioral trends. These pathways are important in considering 
the evolution and development of ant community structure. 
Summary and Conclusions 
We have argued that a very simple hypothesis is sufficient to 
generate predictions of spatial distributions of colonies under a 
variety of ecological settings. The majority of cases in the literature 
(Table 1) support the hypothesis that most ant species are regularly 
distributed with respect to conspecifics and other co-occurring 
species. We assert that this is a natural outcome of high overlap in 
food utilization in many species, and in particular, among general- 
ists. We have suggested that departures from expected spatial 
patterns be used as a measure of competition between species, but 
too little information on colony foraging radii in relation to spacing 
patterns exists to test our hypotheses critically. More field measure- 
ments of colony foraging distances in relation to intercolony spacing 
are needed. Measurements of potential foraging distances when 
