128 
Psyche 
[Vol. 92 
Veromessor pergandei is strongly pleometrotic; 89.1% of all 
queens found in 1983 came from multiple queen associations (Fig. 
2). Formation of associations is not density-dependent. Percentage 
of all queens found pleometrotic during regular surveys of ravines in 
1984 is uncorrelated with queen density for that survey (r = —.058; 
N = 9 days when more than one queen was found). 
While searching for foundress associations we regularly found 
dealate queens in a group under rocks or similar objects in an 
immobile, “pupoid” state (i.e. legs and head held tightly under the 
thorax as seen in ant pupae). “Pupoid” queens were usually found 
when environmental conditions, especially soil temperature, were 
conducive to V. pergandei activity as indicated by simultaneous 
foraging of nearby adult colonies. Exact function of “pupoid” 
behavior, following flight and mating but preceeding excavation, is 
unclear but may promote accumulation of queens into a developing 
pleometrotic association. There appears a clear cost of predation 
associated with any such behavior that delays excavation of a bur- 
row. We frequently observed up to 50% of the starting nests in an 
area destroyed (either by a rodent or lizard); additionally, we 
observed predation on foundresses at or near the soil surface by 
centipedes, spiders and fire ants ( Solenopsis sp.). 
Discussion 
The mating season of V. pergandei is quite long when compared 
with sympatric seed-harvester species. For example, Pogonomyr- 
mex spp. have a brief mating season determined by summer rains; 
colonies release all alates during one or several days of flight activity 
(Holldobler 1976a, Markl et al. 1977, Davidson 1982). By compari- 
son, V. pergandei releases alates over a 2 mo. period; frequently a 
colony releases only a small number of alates for many days (Fig. 1). 
Method of colony foundation also differs dramatically between V. 
pergandei and Pogonomyrmex spp. While V. pergandei frequently 
forms foundress associations (Fig. 2), Pogonomyrmex rugosus 
rarely does so. Of 70 starting P. rugosus colonies excavated in 
Tempe, AZ, 66 contained a single queen (Rissing and Pollock, 
unpublished data). Similarly, Holldobler (1976b) reports that 
“hundreds” of founding P. rugosus, P. barbatus and P. maricopa 
colonies contained only a single queen. 
