Host seeking behaviour of Anopheles mosquitoes in response to olfactory and visual cues

The urgency to develop new tools for malaria control to supplement existing ones (namely long-lasting insecticide-treated bed nets (LLINs) and indoor residual spraying (IRS), has been fuelled by the recent stagnation in control of malaria, especially in sub-Saharan Africa. This slow-down in control, and resurgence in some areas, has been attributed to increasing insecticide resistance in major malaria vectors and changes in mosquito populations from predominately endophilic and anthropophilic to exophilic and opportunistic feeders. The rising importance of outdoor-biting malaria vectors in maintaining residual malaria transmission has warranted the development of control tools that target these species. Due to the mainly indoor protection conferred by LLINs and IRS, development of new control tactics is required to target mosquito populations that bite and rest predominately outdoors. Tools previously developed to target malaria mosquitoes when used singly were experimentally combined towards synergizing their protective efficacy outdoors. Spatial repellents have received renewed interest due to their ability to prevent host-vector contact. Spatial repellents can be used to limit contact with outdoor-biting mosquitoes and in many cases, are also effective against mosquitoes which have been shown to be insecticide resistant. When used in combination with odour-baited traps, spatial repellents present opportunities for mosquito control through the use of a proposed ‘push-pull’ strategy in controlling outdoor-biting mosquitoes. This strategy could be effective in residual transmission settings.

In order to understand the dynamics that a push-pull mosquito control strategy would involve, a number of research questions were studied; 1) Does the inclusion of an optimally-performing spatial repellent in an eave fabric reduce outdoor-biting rates of malaria vectors? 2) Does the addition of an optimised odour-baited trap to the spatial repellent improve on any protection conferred by the spatial repellent alone? 3) Are the airborne concentrations of the chemical components of the spatial repellent and odour trap quantifiable towards determining the effective range of possible protection?

In Chapter 2, I outlined systematic investigations of different concentrations of PMD and transfluthrin as potential spatial repellents applied on eave fabric and their protective efficacy on unprotected persons in the peri-domestic space outdoors. These outcomes were quantified in semi field systems in western Kenya using insectary-reared Anopheles arabiensis. Two concentrations of PMD tested did not offer any protection against outdoor biting, while two concentrations of transfluthrin were both protective, with the higher concentration offering more protection than the lower one. These results confirmed the spatial repellency effect of transfluthrin on outdoor-biting An. arabiensis. Supplementation of transfluthrin using an MB5-baited Suna trap did not have any additional effect on the protective efficacy seen with the spatial repellent alone. The MB5-baited Suna trap supplemented with either carbon dioxide or 2-butanone on its own did not divert mosquitoes from an unprotected human seated 2.5 m away from the trap, confirming that an unprotected human remains more attractive to host-seeking malaria vectors than to a trap baited with a synthetic lure. Utilization of 2-butanone as a supplement to replace carbon dioxide in the MB5-baited Suna trap was not as effective as mosquito recapture rates of a trap supplemented with carbon dioxide were significantly higher, indicating that 2-butanone is not an adequate carbon dioxide mimic in attracting An. arabiensis. Quantification of the airborne concentrations of the push-pull chemical components showed that transfluthrin was detectable within a 5 m radius from the eave fabrics applied on experimental houses, with higher concentrations detected nearest the house and closer to the ground, with reducing concentrations further away and at distances higher above ground.

In Chapter 3, I explored whether exposure of mosquitoes to transfluthrin has an impact on the trapping efficiency of the MB5-baited Suna trap compared to the CDC UV light trap. This study was carried out to provide a basis for considering alternative pull components in the push-pull system. In the absence of transfluthrin, the odour-baited Suna trap caught more An. arabiensis and An. gambiae s.s. than the CDC UV light trap, indicative of the strong and specific attractive lure offered by the odour-baited traps that is absent in the light trap. In the presence of transfluthrin, a significant reduction in the catches was seen in the odour-baited trap, from recapturing over 30% to recapturing less than 5%, which indicated that transfluthrin was associated with reduced mosquito perception of the synthetic lure from the odour-baited trap. Trapping efficiency of the light trap remained fairly constant at approximately 12%, even in the presence of transfluthrin as mosquito perception of light remained unaffected in the presence of spatial repellents. The efficiency of the light trap was generally quite low as it could only recapture a tenth of released mosquitoes compared to recapture rates of odour-baited traps which were at least three times more in the absence of the repellent and as such, use of a light trap as a potential pull component could not be considered as a replacement for an odour-baited trap.

In Chapter 4, I investigated the possible advantages of presenting transfluthrin on an eave fabric that partially covered the eave gap as a strip, compared with use of a screen that requires complete coverage of the eave gap. Transfluthrin-treated eave strips and eave screens provided similar protection against bites both indoors and outdoors, but eave strips would provide a more cost-effective solution, especially when scaled-up, due to the utilisation of less fabric and increasing the ease of application so that the approach could be easily applied by community members without specialised training in house improvement techniques. Additionally, my investigations were able to reiterate the importance of eave screening as a form of house modification as the application of an untreated eave screen significantly reduced indoor-biting mosquitoes.

In Chapter 5, I carried out field evaluations of the push-pull strategy in a village in western Kenya to establish its performance under natural conditions. The application of transfluthrin-treated eave fabric strips alone significantly reduced the number of An. gambiae s.l. and An. funestus s.l. caught indoors using human landing catches (HLC) compared to catches in the presence of an untreated eave fabric strip and an unbaited Suna trap. The addition of an MB5-baited Suna trap supplemented with carbon dioxide did not have an additional effect on indoor capture rates compared with the protection conferred by transfluthrin-alone. Non-malaria mosquito catches were significantly reduced outdoors using HLC, both in the presence of the repellent alone or repellent with odour-baited trap combined. Outdoor catches of primary malaria vectors were not different in the presence of any of the treatments applied compared to catches in the presence of untreated eave strip and unbaited Suna trap.

The final chapter of this thesis provides a general discussion of the observations seen in the various chapters as well as conclusions obtained.

In summary, transfluthrin released from a strip of eave fabric was able to effectively reduce indoor-biting rates of malaria vectors in both artificial and natural conditions with a possible temperature-dependant protection seen outdoors in semi-field systems. Addition of an odour-baited Suna trap to transfluthrin did not have any effect on protection conferred to persons, both indoors and outdoors. Spatial availability of transfluthrin in the airspace around the peri-domestic area was confirmed and would be useful towards generating data for modelling possible area range of protection.  

The push-pull strategy presents a tool that can be developed to reduce outdoor-biting rates by mosquitoes found in East Africa. The strategy may need to be optimised to provide a protective action against infective bites from malaria vectors under alternative geographical and climatic conditions. Presentation of spatial repellents in a strip of eave fabric that does not fully cover the eave gap is just as protective as a repellent-treated eave screen and could present greater opportunity for cost-effective control of malaria in areas where improvement of housing is not immediately possible. Generation of a more effective attract and kill tool that can supplement transfluthrin would improve the overall performance of the push-pull strategy through greater suppression of mosquito populations. In the long run, identification of an even better, non-insecticidal, spatial repellent that is cheap, long-lasting and effective against a range of mosquito species in various geographical conditions would be ideal in ensuring effective disease control.