Behavioural responses of females of two anopheline mosquito species to human-occupied, insecticide-treated and untreated bed nets

BACKGROUND: Insecticide-treated bed nets (ITNs), used extensively to reduce human exposure to malaria, work through physical and chemical means to block or deter host-seeking mosquitoes. Despite the importance of ITNs, very little is known about how host-seeking mosquitoes behave around occupied bed nets. As a result, evidence-based evaluations of the effects of physical damage on bed net effectiveness are not possible and there is a dearth of knowledge on which to base ITN design. METHODS: The dispersion of colony-raised female Anopheles gambiae and Anopheles albimanus was observed in 2-hr laboratory experiments in which up to 200 mosquitoes were released inside a mosquito-proof 3 m × 3 m tent housing a bed net arrayed with 18 30 cm × 30 cm sticky screen squares on the sides, ends and roof. Numbers of mosquitoes caught on the sticky squares were interpreted as the ‘mosquito pressure’ on that part of the net. RESULTS: Presence of a human subject in the bed net significantly increased total mosquito pressure on the net for both species and significantly re-oriented An. gambiae to the roof of the net. Anopheles albimanus pressure was greatest on the bed net roof in both host-present and no-host conditions. The effects of different human subjects in the bed net, of different ambient conditions (dry, cool conditions vs warm, humid conditions) and of bed net treatment (deltamethrin-treated or no insecticide) on mosquito pressure patterns were tested for both species. Species-specific pressure patterns did not vary greatly as a result of any of these factors though some differences were noted that may be due the size of the different human subjects. CONCLUSIONS: As a result of the interaction between host-seeking responses and the convective plume from the net occupant, species-specific mosquito pressure patterns manifest more or less predictably on the bed net. This has implications for bed net design and suggests that current methods of assessing damaged bed nets, which do not take damage location into account, should be modified.