Thermal preferences of subtropical Aedes aegypti and temperate Ae. japonicus mosquitoes

Temperature is an important determinant affecting the capacity of disease vectors like mosquitoes (Culicidae) to transmit disease agents. Although the impact of temperature on vector-borne disease dynamics has been studied intensively, the actual temperature encountered by the vector in a heterogeneous landscape is rarely taken into account. If disease vectors have temperature preferences and therefore select specific microhabitats, this would substantially influence key life history traits that determine transmission intensity. The thermal preferences of subtropical Aedes aegypti and temperate Ae. japonicus mosquitoes were investigated in a temperature gradient setup consisting of a Plexiglas box on top of an aluminium plate on two thermal regulators. Blood-fed (one day after feeding) and unfed (non-blood-fed) mosquitoes were released in small (15-20 degrees C, 20-25 degrees C, 25-30 degrees C) and large (15-30 degrees C, 30-45 degrees C) temperature gradients to assess their thermal preferences after 15 min. Additionally, the effect of humidity was investigated in a two-choice chamber setup. Both mosquito species avoided higher temperatures, pronouncedly dangerously high temperatures in the 30-45 degrees C gradient. At lower temperatures, blood-fed mosquitoes preferred the cooler sides of the 20-25 degrees C and 25-30 degrees C gradient, which were all below their rearing temperature. In the lowest gradient of 15-20 degrees C, no preferences were found. The thermal preference of unfed mosquitoes was similar to that of the blood-fed mosquitoes. No humidity preference or effect of humidity on temperature preferences was observed within the tested range (40-90%). The set-up allows for assessing the thermal preference of mosquitoes under controlled conditions. The observed preference of mosquitoes for cooler temperatures would increase their longevity and slow down pathogen development. If similar micmhabitat selection is observed in the field, vector borne disease models should be adjusted accordingly.