SEX-RATIO SELECTION IN A BIVOLTINE THRIPS. I. CONDITIONAL SEX-RATIO MANIPULATION AND FITNESS VARIATION

Females of the bivoltine thrips Elaphrothrips tuberculatus (Hood) (Insecta: Thysanoptera) produce broods of either all males (by viviparity) or all females (by oviparity). Measurements of the sex-allocation ratio, ecological and physiological conditions affecting male and female offspring body size, and correlates of the relative fitnesses of adult males and females in relation to size indicate that female parents tend to be viviparous (produce males) if their offspring will become relatively large adults, and that males gain more in fitness from large size than do females. However, the conditions that link sex allocation with offspring fitness differ between the spring and summer generations. In spring, when breeding is synchronous, 1) oviparous and viviparous females do not differ in body size, 2) females tend to be viviparous where the fungus upon which they feed is relatively dense and where their offspring will become relatively large adults, and 3) fungus density is highly correlated with male and female offspring size. In summer, when breeding is relatively asynchronous, 1) viviparous females are much larger than oviparous females early (but not late) in the season, 2) large viviparous females begin breeding earlier than smaller ones, 3) offspring developing earlier in the season become larger adults, and 4) a higher proportion of females are viviparous earlier than later. Field experiments and field collections show that the covariation among sex allocation, conditions, and fitness is not caused by differential mortality by size or sex. Differences between the spring and summer generations in the cues used by females to adjust offspring sex ratio may be caused by seasonal variation in the factors that affect offspring size. However, in both generations, females tend to produce sons only when their offspring will become relatively large adults, whereas daughters are produced regardless of offspring size. These data suggest that females of E. tuberculatus avoid production of males (the sex with higher variance in expected fitness) when the size of their offspring is relatively uncertain.