Taxis assays measure directional movement of mosquitoes to olfactory cues

Abstract

Background: Malaria control methods targeting indoor-biting mosquitoes have limited impact on vectors that feed and rest outdoors. Exploiting mosquito olfactory behaviour to reduce blood-feeding outdoors might be a sustainable approach to complement existing control strategies. Methodologies that can objectively quantify responses to odour under realistic field conditions and allow high-throughput screening of many compounds are required for development of effective odour-based control strategies.

Methods: The olfactory responses of laboratory-reared Anopheles gambiae in a semi-field tunnel and A. arabiensis females in an outdoor field setting to three stimuli, namely whole human odour, a synthetic blend of carboxylic acids plus carbon dioxide and CO(2) alone at four distances up to 100 metres were measured in two experiments using three-chambered taxis boxes that allow mosquito responses to natural or experimentally-introduced odour cues to be quantified.

Results: Taxis box assays could detect both activation of flight and directional mosquito movement. Significantly more (6-18%) A. arabiensis mosquitoes were attracted to natural human odour in the field up to 30 metres compared to controls, and blended synthetic human odours attracted 20% more A. gambiae in the semi-field tunnel up to 70 metres. Whereas CO(2) elicited no response in A. arabiensis in the open field, it was attractive to A. gambiae up to 50 metres (65% attraction compared to 36% in controls).

Conclusions: We have developed a simple reproducible system to allow for the comparison of compounds that are active over medium- to long-ranges in semi-field or full-field environments. Knowing the natural range of attraction of anopheline mosquitoes to potential blood sources has substantial implications for the design of malaria control strategies, and adds to the understanding of olfactory behaviour in mosquitoes. This experimental strategy could also be extended from malaria vectors to other motile arthropods of medical, veterinary and agricultural significance.

Figure 1

Experimental set-up. A) Closed taxis box, and B) taxis box in an open position. Mosquitoes were placed in the middle chamber and remained there until a pulley mechanism was used to lift the barriers, allowing access to the two side chambers. C) Due to the design of the entry points, mosquitoes could easily fly into a side chamber, but could not return to the centre. Taxis boxes containing mosquitoes exhibiting D1) positive taxis (majority move towards (T) the stimulus), D2) kinesis (mosquitoes distributed randomly between the three chambers), and D3) negative taxis (majority move away (A) from the stimulus) in response to a directional olfactory stimulus indicated by the arrow. E) Taxis boxes within the semi-field tunnel, and F) four taxis boxes set up in parallel at each distance for the full open field experiment in Ifakara, Tanzania. The human experimenter opens and closes the barriers separating the three chambers by a pulley mechanism to allow mosquitoes to orient towards or away from the point of stimulation.

Figure 2

Activation (A) and taxis movement (B) of A. gambiae s.s. towards stimuli in the semi-field tunnel. Model-estimated proportion of A. gambiae s.s. mosquitoes A) moving out of the middle chamber either towards or away from the point of stimulation (= activation) or B) moving towards the point of stimulation (= taxis) by distance for control nights, i.e. no stimulus (filled circle), 500 ml min^-1 CO₂ (open circle), synthetic odour blend + 500 ml min^-1 CO₂ (open square) and a human volunteer (open diamond) in the semi-field tunnel (Experiment 1). Each point represents the model-estimated proportion from the best-fitting model, with the associated 95% confidence intervals [59] indicated by vertical bars. NB. At each distance, the data points for the four treatments are slightly offset from one another to improve clarity.

Figure 3

Activation (A) and taxis movement (B) of A. gambiae s.s. towards stimuli in the semi-field tunnel. Model-estimated proportion of A. arabiensis mosquitoes A) moving out of the middle chamber either towards or away from the point of stimulation (= activation) or B) moving towards the point of stimulation (= taxis) by distance for control nights, i.e. no stimulus (filled circle), 500 ml min^-1 CO₂ (open circle), synthetic odour blend + 500 ml min^-1 CO₂ (open square) and a human volunteer (open diamond) in the open field environment (Experiment 2). Each point represents the model –estimated proportion from the best-fitting model, with the associated 95% confidence intervals [57] indicated by vertical bars. NB. At each distance, the data points for the four treatments are slightly offset from one another to improve clarity.