Published online by Cambridge University Press: 10 July 2009
A rickettsia-free (aposymbiotic) stock of Aedes polynesiensis Marks (POLY-A) was crossed with (a) three symbiont-infected stocks of A. polynesiensis (POLY-S from Samoa and POLY-N and POLY-T from Fiji), (b) A. pseudoscutellaris (Theo.) (PSE) from Fiji, (c) A. alcasidi Huang (ALC), A. scutellaris katherinensis Woodhill (KATH), A. s. malayensis Colless (MAL) and A. s. scutellaris (Wlk.) (SCUT), which occur to the west of Fiji and (d) an aposymbiotic stock of A. cooki Belkin (CO) from Niue. It was bidirectionally compatible with CO, but in all other crosses compatibility was high when POLY-A was the male parent and very low when it was the female parent. Backcross data suggested that the crossing type was maternally inherited. ALC and KATH were bidirectionally compatible; both were virtually incompatible with POLY-S and PSE, compatible with SCUT when the latter was the female parent, and compatible with CO when the latter was the male parent. POLY-S females were moderately compatible with a third Fijian stock of A. polynesiensis (POLY-V), and POLY-N and POLY-V were compatible with PSE. If, as in Culex pipiens L., rickettsia-like symbionts are responsible for cytoplasmic incompatibility, then aposymbiotic males should cross successfully with symbiont-infected females, whereas the reciprocal cross should be unsuccessful. Since PSE, MAL and SCUT contain symbionts, their crossing relationships are consistent with the hypothesis. However, ALC and KATH appear to be aposymbiotic and their crossing relationships conflict with the hypothesis. There is little evidence of behavioural barriers to mating between species, but whereas male hybrids of two eastern species were capable of normal insemination, male hybrids between western and eastern species gave very low insemination rates. The egg-hatch rates from backcrosses of female hybrids between western and eastern species to the parents were reduced.