Female multiple mating (polyandry) is widespread across Insecta, even if mating can be costly to females. To explain the evolution and maintenance of polyandry, several hypotheses, mainly focusing on the material (direct) and/or the genetic (indirect) benefits, have been proposed and empirically tested in many species. Considering only the direct benefits, repeatedly-mated females are expected to exhibit the same fitness as multiply-mated females under the same mating frequency. In the present study, we compare the fitness of females received monandrous repeated mating (MM) and polyandrous multiple mating (PM) in a polyandrous leaf beetle Galerucella birmanica and assess female mate preference with regard to polyandry or monandry. Our data indicate that the longevity and the egg-laying duration of MM females are significantly longer than that of PM females. MM females produce significantly more hatched eggs than PM females over their lifetime under the same mating frequency, which results from the high hatching rate of eggs produced by MM females. PM females mated with novel virgin males in the second mating suffer decreased longevity and lifetime fecundity compared with PM females mated with novel mated males in the second mating. Once-mated females are more likely to re-mate with familiar males than novel males. By contrast to expectations, the results of the present study suggest that repeated mating provides females with more direct benefits than multiple mating in G. birmanica, and females prefer to re-mate with familiar males. The possible causes of this finding are discussed.

The cotton-melon aphid, Aphis gossypii Glover, is a polyphagous insect pest with many host-specialized biotypes, such as the Cucurbitaceae- and Malvaceae-specialized (CU and MA) biotypes. Bacterial symbionts were reported to determine the host range in some aphids. Whether this is the case in A. gossypii remains unknown. Here, we tested the host specificity of the CU and MA biotypes, compared the host specificity between the wingless and winged morph within the same biotype, and analyzed the composition of the bacterial symbionts. The reproduction of the CU and MA biotypes reduced by 66.67% and 82.79%, respectively, on non-native hosts, compared with on native hosts. The composition of bacterial symbionts was not significantly different between the CU and MA biotypes, with a Buchnera abundance >95% in both biotypes. Meanwhile, the winged morph produced significantly more nymphs than the wingless morph on non-native hosts, and the Buchnera abundance in the winged morph was only about 10% of that in the wingless morph. There seemed to be a relationship between the Buchnera abundance and host specificity. We regulated the Buchnera abundance by temperature and antibiotics, but did not find that a low Buchnera abundance resulted in the high reproduction on non-native hosts. We conclude that the host specificity of A. gossypii is not controlled by specific bacterial symbionts or by Buchnera abundance.