Sociality in <I>Amphylaeus morosus</I> (Hymenoptera: Colletidae: Hylaeinae)

Abstract

To date, social nesting has been relatively unknown in the bee family Colletidae. Same-generation females of the bee Amphylaeus morosus Smith frequently share tubular nests where only 1 brood cell can be provisioned at a time. In montane eucalypt forests of southern Australia, A. morosus nests are constructed within naturally excised fronds of the rough tree fern, Cyathea australis. In these habitats, the species is univoltine with adult eclosion occurring in late summer. Dispersal from overwintering nests and new nest initiation begins in spring, and all new nests contain only 1 adult female. However, a significant proportion of old, reused nests during the brood-rearing phase contain 2 or 3 adult females that are mated with mature ovaries (≈23% of all inhabited nests). Per capita brood production was statistically higher in reused versus new nests, probably because the former start their cell provisioning earlier. Per capita brood production was not statistically higher in two- versus one-female reused nests, although there were trends in this direction and earlier adult female mortality may have obscured end-of-season patterns. Brood parasitization by a gasteruptiid wasp (Gasteruption sp.) was statistically higher in new nests than two-female reused nests, but only marginally higher in one- versus two-female reused nests. Average relatedness among adult females in multifemale nests was quite low (r = 0.26 ± 0.06 SE). Hence, kin selection is unlikely to be a major factor selecting for cooperative nesting in this species. Cell provisioning patterns in A. morosus could lead to nestmates conflicting over foraging effort and reproduction. No evidence for reproductive division of labor or foraging specialization among nestmates was found, however, and their relative body sizes, ovary sizes, and wing wear were not significantly different from statistical expectation. Therefore, sociality in A. morosus probably results from the benefits of sharing a valuable resource (i.e., a preexisting nest) and avoiding dispersal costs, rather than cooperation per se. Nevertheless, the prospects for nestmate cheating are high. High-resolution genetic studies could determine if the apparent lack of reproductive skew, based on dissection data, reflects brood maternity. Sociality in A. morosus provides an ideal opportunity to investigate reproductive skew theories.