218 
Psyche 
[Vol. 93 
a complete understanding of population fluidity must be predicated 
on work conducted in the field, our results provide insight into the 
evolutionary ecology of this ant. 
Population structure in L. longispinosus 
Many ants of the genus Leptothorax are polygynous (Buschinger 
1968, 1974), and L. longispinosus is no exception (Talbot 1957; 
Headley 1943; Alloway et al 1982). Previous work on the E. N. 
Huyck Preserve (Albany County, New York) showed the popula- 
tion to be facultatively polygynous: some nests contain no queen, 
others have one, and still others have multiple queens (Herbers 
1984) . Moreover, there was a strong winter-summer dichotomy in 
queen distribution. Many nests in summer are queenless, whereas in 
winter such groups are rare (Herbers 1986a); similarly, the average 
number of queens per nest is lower in summer. Finally, nests are 
considerably more spread out in summer than in winter (Herbers 
1985) . These results are best explained as correlates of a seasonal 
shift in spatial structure: colony fractionation in summer and con- 
densation for overwintering. It appears that, for the most part, 
overwintering nests are independent colonies that become poly- 
domous in summer when they fractionate to occupy several nest 
sites (Herbers 1986a). This cyclic polydomy hypothesis is supported 
by behavioral evidence reported below. 
Methods 
Nests of L. longispinosus were excavated from the New York site 
in late October 1983, when they exhibited spatial relationships and a 
distribution of queens among nests that is typical of winter. Each 
nest was returned to the laboratory and removed from its stick, 
acorn, or root. The ants were then resettled into glass tubes 10 cm 
long and 4 mm in diameter. Each nest was put into a separate box 
and incubated at 4°C for overwintering. In March the temperature 
and light-dark cycles were slowly incremented to match outside 
conditions. On May 8, 1984 the conditions were stabilized at 14 
hours of light. On that date, we positioned 17 nests on 4 artificial 
forest floors to duplicate their spatial positions in nature the pre- 
vious fall (Figure 1). Observations and censuses were then con- 
ducted until August 27, 1984, when the experiments were terminated. 
