Background: Holes in netting provide potential routes for mosquitoes to enter ITNs. Despite this, there is little information on how mosquitoes respond to holes in bed nets and how their responses are affected by hole size, shape and orientation or by ambient conditions around the net.
Methods: Female Anopheles gambiae (G3) were recorded in a simulated bed net consisting of two sizes of untreated netting-covered behavioural arenas placed above and beside (to simulate the bed net roof and sides respectively) the experimenter who was a source of host cues from 'inside' the net. A round hole of 9 mm or 13 mm diameter was cut into the centre of the netting of each arena. Videos of unfed female mosquitoes in arenas were analysed for time spent flying, walking and standing still and for exit through the hole. The effects of the experimenter on temperature and relative humidity around the simulated net were also measured.
Results: Mosquitoes were significantly more active in overhead arenas than in arenas to the side. Hole passage was significantly more likely in smaller arenas than larger ones and for larger holes than smaller ones. In arenas to the side, hole passage rate through small holes was about 50% less likely than what could be explained by area alone. Passage rate through holes in overhead arenas was consistent with hole area. Temperature in arenas did not strongly reflect the experimenter's presence in the simulated net. Relative humidity and absolute humidity in overhead arenas, but not in arenas to the side, were immediately affected by experimenter presence.
Conclusions: Higher levels of activity in overhead arenas than in arenas to the side were likely due to the rising heat and humidity plume from the experimenter. Lower than expected passage rates through smaller vertically oriented holes may have been be due to an edge effect that does not apply to horizontally oriented holes. Results suggest that current methods of assessing the importance of physical damage to ITNs may not accurately reflect mosquito entry risk in all cases.