In a colony headed by a single monandrous foundress, theories predict that conflicts between a queen and her workers over both sex ratio and male production should be intense. If production of males by workers is a function of colony size, this should affect sex ratios, but few studies have examined how queens and workers resolve both conflicts simultaneously. We conducted field and laboratory studies to test whether sex-ratio variation can be explained by conflict over male production between queen and workers in the primitively eusocial wasp Polistes chinensis antennalis. Worker oviposition rate increased more rapidly with colony size than did queen oviposition. Allozyme and microsatellite markers revealed that the mean frequency of workers' sons among male adults in queen-right colonies was 0.39 +/- 0.08 SE (n = 22). Genetic relatedness among female nestmates was high (0.654-0.796), showing that colonies usually had a single, monandrous queen. The mean sex allocation ratio (male investment/male and gyne investments) of 46 queen-right colonies was 0.47 +/- 0.02, and for 25 orphaned colonies was 0.86 +/- 0.04. The observed sex allocation ratio was likely to be under queen control. For queen-right colonies, the larger colonies invested more in males and produced reproductives protandrously and/or simultaneously, whereas the smaller colonies invested more in females and produced reproductives protogynously. Instead of positive relationships between colony size and worker oviposition rate, the frequency of workers' sons within queen-right colonies did not increase with colony size. These results suggest that queens control colony investment, even though they allow worker oviposition in queen-right colonies. Eggs laid by workers may be policed by the queen and/or fellow workers. Worker oviposition did not influence the outcome of sex allocation ratio as a straightforward function of colony size.

Beyond helping the queen's reproductive function, the worker caste of Polistes wasps is expected to have an alternative reproductive strategy; to become a replacement queen, following copulation with early males that emerge along with the first brood workers. We analyzed genotypes and chromosome numbers of the early males in a natural population of P. chinensis antennalis. The frequency of earlymale-producing colonies within a population was 5.2 % (4 /77) in 1993 through 1995. The numerical sex ratios in the 4 colonies for the 3 years and another early male-producing colony in 1999 were 1:1. For one early-male producing colony in 1995, we genotyped 31 females and all 128 males. All 31 genotyped females out of 67 emerged females showed heterozygosity and 69 out of 128 genotyped males were heterozygous; the remaining 59 males were homozygous or hemizygous. The ratio between the number of females and diploid males in this colony was 67 :69, which is almost identical with one-to-one ratio.The cytological analysis of male chromosome for one early-male producing colony in 1999 revealed that the early males laid by a queen had 46 chromosomes, while worker laid males had 23. Our results revealed that the early males in this species are diploid males and eventually sterile. Secondly, the single locus complementary sex determination model is most plausible to explain the diploid male production in this species.