Indoor and outdoor biting behaviour of malaria vectors and the potential risk factors that enhance malaria in southern Malawi

Current methods of malaria vector control implemented by national control programmes rely mainly on the use of insecticides. These include the use of long-lasting insecticide treated nets (LLINs) and indoor residual spraying (IRS). The success of LLINs and IRS is underpinned by the protection from infectious mosquito bites provided to individuals and the reduction in mosquito population size caused by sufficient contact of mosquitoes with the insecticide in the nets or on the walls of the houses. The high degree of endophily (resting indoors) and endophagy (feeding indoors) exhibited by the dominant African malaria vectors has been, therefore, a key component of that success. However, in recent years in some regions, the biting behaviour of the African malaria vectors, both indoors and outdoors and during a wider range of times than previously recognized, has been reported. This has an implication on malaria control because individuals are at risk of receiving infectious bites from vectors that are biting either outdoors or indoors at times when people are not protected by the primary control tools. Additionally, resistance of mosquitoes to these insecticides exacerbates the risk for malaria transmission. Therefore, understanding the degree of endophagy/exophagy of the vectors, when or where humans are exposed to mosquito bites, entry points for malaria vectors into houses and biological factors enhancing malaria transmission in a region is important. The collective information from studying these natural behavioural aspects of mosquitoes will help in designing interventions that protect individuals from infective mosquito bites, thereby reducing malaria transmission and disease burden. The research described in this thesis focused on the biting behaviour of malaria vectors in and around houses in southern Malawi. Chapter 2 provides an overview of the biting times of malaria vectors in Africa, both historically and currently. Our literature search showed that the biting behaviour of mosquitoes both indoors and outdoors was common but the biting peaks vary across and within regions. We explored the factors that may be associated with the variations in the biting behaviour of the vectors. We found that the availability of hosts is one of the potential factors. Furthermore, there is a likelihood that the prolonged use of LLINs may lead to variations in the biting behaviour of malaria vectors although in some regions where such variations have been reported, they rely on data after the implementation of LLINs only. In Chapter 3, the biting patterns of mosquitoes were assessed both indoors and outdoors and during the wet and dry seasons. We found that the major malaria vectors were Anopheles arabiensis and An. funestus. Whereas An. arabiensis was more likely to bite outdoors than indoors, An. funestus was more likely to bite indoors than outdoors. During the dry season, the biting activity of An. gambiae s.l. was constant outdoors across the time of observation (18:00 h to 08:45 h), but highest in the late evening hours (21:00 h to 23:45 h) during the wet season. The biting activity of An. funestus s.l. was highest in the late evening hours (21:00 h to 23:45 h) during the dry season and in the late night hours (03:00 h to 05:45 h) during the wet season. Biting activities that occurred in the late evening hours, both indoors and outdoors, coincided with the times at which individuals may still be awake and physically active, and therefore unprotected by LLINs. Additionally, a substantial number of anopheline bites occurred outdoors. These findings imply that LLINs would only provide partial protection from malaria vectors, which would affect malaria transmission in this area. Therefore, protection against bites by malaria mosquitoes in the early and late evening hours is essential and can be achieved by designing interventions that reduce vector-host contacts during this period. Results of Chapter 3 highlight the need for effective tools for sampling mosquitoes indoors and outdoors. Chapter 4 compares the efficiency of the Suna trap, an odour baited trap, to that of the human landing catch (HLC) and Centers for Disease Control Light Trap (CDC-LT). We found that use of the Suna trap both indoors and outdoors compares well with that of the HLC. This implies that since the HLC method is labour intensive and expensive at large scale implementations, the Suna trap can serve as a substitute for the HLC for estimating the biting rates. On the other hand, the mosquito catches with the Suna trap were lower than those of the CDC–LT. The effectiveness of the Suna trap in sampling mosquitoes when placed either indoors, outdoors or simultaneously indoors and outdoors is the same (Chapter 4). This finding makes the Suna trap more efficient than the CDC-LT because the use of the latter trap is seemingly dependent on the attractiveness of the individual sleeping under the adjacent bed net and the use of this trap outdoors yields fewer malaria vectors. Additionally, the Suna trap uses synthetic odour baits and does not rely on use of humans as baits as with the HLC or CDC-LT methods. Biological factors such as the presence of cattle around houses has been associated with either a protective effect against bites by vectors as these vectors are diverted to other blood meal hosts such as cattle rather than humans or with more bites as the vectors have sufficient blood meal hosts (humans and cattle). Therefore, Chapter 5 describes the results from an assessment of the impact of cattle on the resting behaviour of malaria vectors. The presence of cattle near a house significantly reduced the abundance of indoor resting An. funestus but not An. arabiensis. This implies that the reduction of the former species was possibly due to the deterrent effect of cow odours. These data suggest that repellents around a house disrupt the host-seeking behaviour of malaria vectors. When combined with attractant traps, the resulting push-pull system would lead to reduction of malaria vectors and hence, malaria transmission. In Chapter 6 the impact of fully and partially closed eaves on house entry rates mosquitoes was studied. We compared mosquitoes in houses with fully closed eaves, open eaves and three levels of partially closed eaves. It was found that fully closed eaves and houses with one small opening on the eave significantly reduced house entry of malaria vectors compared to partially and fully open eaves. The mosquitoes were able to locate the remaining entry points on the eaves, a finding which has an implication on malaria transmission. Therefore, quality control is an important component when implementing structural house improvements. The results of Chapter 3 showed that An. funestus s.s. was more likely to bite indoors than outdoors in the present study region and therefore, a house improvement strategy that includes the closure of eaves may be a complementary tool for vector control that protects against biting by An. funestus which is responsible for the indoor biting in southern Malawi (Chapter 3). In Chapter 7, the general discussion interprets the key findings and links these to the implications for malaria control. Furthermore, the findings described in this research provide recommendations for future research. It is concluded that in southern Malawi, the major malaria vectors are An. arabiensis and An. funestus contributing to outdoor and indoor malaria transmission, respectively. Development of tools that can target the biting activity of these vectors both indoors and outdoors at times when individuals are not under bet nets is highly recommended. Furthermore, a house improvement strategy that includes closure of eaves can significantly reduce house entry by malaria vectors. Additionally, the use of odour-baited traps looks promising as tools for sampling malaria vectors both indoors and outdoors as well as tools for mass trapping of mosquitoes to reduce malaria vectors thereby reducing malaria transmission and burden