|| Checking for direct PDF access through Ovid
Trivers' and Willard's hypothesis that natural selection favors sex allocation in relation to maternal condition assumes iteropary. Though this assumption is not met in most solitary Aculeata, the reproductive life span of semelparous females may be divided into discrete successive cycles by the risk of open-cell parasitism. Females can avoid losing their investment to parasites attacking the open cell only by limiting the provision time for each cell. The restriction of time available for the investment in a single progeny irrespective of the condition of the female leads to de facto iteropary. Moreover, in Hymenoptera, there are no costs for sex allocation due to the haplodiploid mode of sex determination. In sexually size dimorphic species, females in poor condition are predicted to invest in the smaller sex and vice versa. The resulting prediction of a conditional sex allocation in solitary Aculeata was tested in the Red Mason bee, Osmia rufa (Osmia bicornis), a stem or hole-nesting, polylectic, univoltine megachilid bee. Body size is a key component of condition in females of nest-constructing solitary bees. Large females collect the same amount of pollen and nectar in a shorter time than small ones and should suffer less from parasitism. We found that small females dealt with their handicap of a low provisioning performance by shifting the sex ratio toward sons (the smaller sex) and by reducing the body size of daughters. Large females, however, shifted their offspring sex ratio toward daughters, the sex that depends more on body size in its reproductive value. The sex ratio in the population met the expected Fisherian sex ratio. Although females allocated their investment in the sexes according to their body mass, the population-level investment was balanced.