1985] 
Pollock & Rissing — Veromessor 
131 
involved. We have searched for V. pergandei mating sites during the 
last 3 yr without success. Several observations, however, suggest 
such sites are small, local and on or near the ground. We have found 
horned lizard ( Phrynosoma sp.) fecal pellets consisting soley of V. 
pergandei alate exoskeletons at South Mountain. Absence of 
workers in these fecal pellets suggests alates were not consumed at 
the nest site, where workers would have vigorously mobbed any 
horned lizard (Rissing 1981). We have also noticed starting colonies 
are clumped along ravines at South Mountain even when queen 
density is low. Live males occur frequently in these clumps, under 
debris within 1 m of starting nests or actually associated with one or 
several females. These observations, along with the high percentage 
of fertilized females found throughout the protracted mating sea- 
son, suggest mating occurs on or near the ground in small, localized 
groups. These observations also suggest males and females do not 
differentially disperse after mating; this is unusual among insects 
(Thornhill and Alcock 1983). 
Another line of evidence supports this view of V. pergandei ’s 
mating system. If alates mate in small, localized groups, males do 
not compete panmitically for females. Such limited alate dispersal 
should select for female biased ratios of investment among parental 
colonies (Hamilton 1967, Bulmer and Taylor 1980, Taylor and 
Bulmer 1980, Charnov 1982, Wilson 1983). All of the colonies we 
have excavated had such a bias or contained only males (Table 1). 
This investment pattern cannot be explained through worker control 
of sex ratio, for observed biases were three times as large as those 
predicted under worker control (Trivers and Hare 1976, Benford 
1978, Oster and Wilson 1978, Charnov 1982). If our sampling 
method is an adequate measure of a colony’s partitioning of sexual 
investment, this is strong evidence that V. pergandei alates exhibit 
low dispersal with respect to their parental colony-mates. This is 
what we would expect when colonies release alates asynchronously 
over a lengthy period. 
Veromessor pergandei is sympatric with several Pogonomyrmex 
spp. throughout its range and occurs exclusively in more xeric areas 
of the Mohave Desert (Creighton 1950, Wheeler and Wheeler 1973, 
Rissing 1981). If our analysis of V. pergandei’s mating system is 
correct, it offers a strong contrast to that of Pogonomyrmex spp. 
Pogonomyrmex produce a few massive mating aggregations (“leks”), 
