The vector competences of 6 geographic strains of Aedes polynesiensis for Wuchereria bancrofti were studied using two types of experimental infections. Experimental infection of laboratory-bred mosquitoes fed on the carriers' forearms with different levels of microfilaraemia showed that microfilariae (mf) uptake was directly proportional to the carrier's mf density and, as mf densities decreased, concentration capacity of Ae. polynesiensis increased. It was also shown that infection has an important effect on mosquito mortality, and that the mortality rate differed among mosquito strains. In infections using artificial feeders, the mf uptake was closely regulated, thus showing differences in the vectorial efficiency of Ae. polynesiensis related to the geographic origin of the mosquito strain. The mosquitoes from the Society archipelago were more efficient intermediate hosts than geographically distant strains when infected with W. bancrofti from an island within the archipelago (Tahiti). Mosquito strains from the Society archipelago developed the highest proportion of infective-stage larvae and exhibited the lowest mortality rate when infected with sympatric Tahitian W. bancrofti.

Developmental rates for Wuchereria bancrofti larvae maturing in the vector Aedes polynesiensis were estimated by analysing stage-frequency data consisting of counts of larval stages in mosquitoes reared at 20, 22·5, 25, 27·5, 30 and 32 °C. Base temperatures (i.e. low temperature thresholds) for W. bancrofti development were estimated by the x-intercept method and the model of Lactin et al. (1995). Resulting values were similar with both methods and were ≈12·5 °C for microfilariae (mf) in thorax, ≈17 °C for L1, 15·5 °C for L2 and 16·5 °C for L3. Upper thresholds estimated by the Lactin et al. model were 29·3 °C for mf, 29·1 °C for L1, 32·2 °C for L2 and 31·5 °C for L3. In addition, an original method was devised for computing the L3 upper threshold, by modelling L3 length shrinkage with temperature. It gave a value of 31·4 °C. At 32 °C, L2 and L3 stages exhibited altered morphology, larvae being shorter and wider than expected. The model of Lactin et al. described adequately the non-linear relationship between developmental rate and temperature, but a linear degree-day approach may be applied for temperatures below 27–28 °C.