Olfactory mediation of oviposition in Anopheles mosquitoes: A new strategy for malaria vector control

Mosquito-borne diseases are among the most serious infectious diseases in the world caused by parasites and pathogens. The African mosquito Anopheles gambiae transmits malaria parasites as well as filarial worms. The mosquitoes feed on humans inside and outside houses, which complicates control strategies. Current control strategies focus mostly on reducing contact with host-seeking mosquitoes through the use of long-lasting insecticide-impregnated nets (LLINs) and indoor residual spraying (IRS). Despite the fact that these tools have lowered the burden of malaria in many parts of the world, transmission is sustained by vectors that bite outdoors. This thesis proposes a new strategy for malaria vector control to be added to existing control tools. It goes beyond blood-feeding processes and it targets all mosquitoes which are feeding indoors and outdoors. The proposed strategy is based on manipulation of oviposition behaviour using chemical cues that mimic olfactory cues emitted by breeding sites. Mosquitoes are attracted to lay their eggs on selected water bodies with specific biotic and abiotic factors. Organic chemical cues, emitted by these sites, are presumed to act as infochemicals enabling these mosquitoes to find suitable breeding sites. Identification of the odours that mediate the site-seeking behaviour of mosquitoes can lead to the development of novel intervention methods for mosquito-borne disease control.

My study marked an important progress in the development of mosquito surveillance and control strategies using infochemicals by: (1) generating knowledge on the biology and ecology of immature stages of Anopheles mosquitoes in the study area, (2) providing evidence that mosquitoes select breeding sites preferentially and lay eggs on suitable sites while avoiding unsuitable ones, (3) identifying and characterizing volatile organic chemicals that mediate oviposition behaviour in mosquitoes, (4) generating evidence that oviposition infochemicals can be used to divert oviposition activities to specific breeding sites, which can be targeted by an environmentally safe larvicide: a lure-and-kill method.

The research aimed to understand the reasons that prevent early and late larval stages from sharing breeding sites, and also why breeding sites with late stage larvae are not immediately succeeded by early stage larvae (Chapter 5). Several studies have indicated a possible role of habitat materials in the production of volatiles chemicals which either attract or deter egg-laying mosquitoes. A review of recently published studies showed that oviposition attractants and deterrents are produced by biotic factors present in breeding sites (Chapter 2). In my study it was found that breeding sites which are occupied by larvae play an important role in the production of odours that mediate oviposition behaviour of mosquitoes (Chapter 3). These infochemicals can potentially be used for manipulation of mosquito behaviour by making protected resources unsuitable for mosquitoes, while luring them towards attractive sources. Such push and pull strategies using ovipostion infochemicals for other mosquitoes were reviewed in order to develop similar surveillance and control strategies for malaria vectors. Among other larviciding products reviewed, bacterial products such as Bacillus sphaericus (Bs) and Bacillus thuringiensis israelensis (Bti) emerged with the potential to be combined with attractive infochemicals in the development of a lure and kill strategy (Chapter 2).

To understand the factors influencing oviposition behaviour of Anopheles mosquitoes, a series of dual-choice behavioural experiments were conducted in the laboratory using larval habitats materials. Results show that the presence of larvae in breeding sites mediates oviposition behaviour of Anopheles mosquitoes (Chapter 3). The presence of early-stage larvae in oviposition cups attracted conspecific gravid mosquitoes to lay eggs, whereas the presence of late-stage larvae in the cups deterred conspecific gravid mosquitoes to lay eggs. The response was density dependent: as larval density increased so did the response, until reaching a threshold. Moreover, this behaviour was manifested more profoundly in the presence of larvae than larval-conditioned water. To test whether larval habitats produce volatiles that mediate oviposition behaviour, the oviposition cups containing larvae were covered by a filter paper to prevent gravid females from seeing the larvae in the cups. Results show that both olfactory and visual cues are involved in mediating the oviposition behaviour of anopheline mosquitoes.

To confirm the findings that larval habitats emit infochemicals that mediate oviposition behaviour of  gravid mosquitoes, larval habitats containing early- or late-stage larvae were placed in containers in the laboratory. Headspace volatiles of the larval microcosm were entrained and analyzed by gas chromatography - mass spectrometry (GC-MS). Analysis of the volatiles emitted by larvae in-vitro led to the identification of 16 compounds, and four of them, when tested with An. gambiae, had an effect on oviposition behaviour (Chapter 4). Volatile chemicals extracted from headspace collections of larvae mediated similar behaviour; volatiles entrapped from young larvae attracted conspecific gravid mosquitoes, whereas those from late-stage larvae deterred conspecific gravid mosquitoes to lay eggs. Nonane and 2,4-pentenedione, which were isolated from headspace collection of larvae, enhanced the attractiveness of a synthetic solution and could be used to increase mosquito egg deposition for monitoring or vector control purposes. Dimethyldisulphide (DMDS) and dimethyltrisulphide (DMTS), which were isolated from headspace collection of late-stage larvae only, decreased mosquito egg deposition in a synthetic solution and may act as a spatial deterrent. Laboratory and semi-field experiments in Tanzania showed similar results and suggested that nonane and 2,4-pentanedione can be used as mid-range cues for oviposition.

The study described in Chapter 5 provided ecological characteristics of the field site where observations on the behaviour and distribution pattern of larvae were done. This is the field site where initial observations on succession of larvae in breeding sites and the actual testing of infochemicals took place. The factors influencing the selection of breeding sites by gravid females and distribution patterns of mosquito larvae were explored. Human activities such as brick-making and rice production were found to influence oviposition site selection and hence constitute a risk factor for malaria and other mosquito-borne diseases. The presence of larvae of Anopheles mosquitoes in breeding sites was associated with vegetation, presence of culicine larvae, shallowness and sunlight. The association between anopheline and culicine larvae is a vital finding that suggests that gravid mosquitoes of the two species use similar cues in selecting suitable breeding sites. Moreover, results from the study suggested that productivity of anopheline larvae in the field can be reduced by introducing larval predators, water management, or water pollution with rotting vegetation. The findings obtained after careful observation of the field site helped to design surveillance and larval control strategies using synthetic infochemicals identified in Chapter 4.

Field experiments in east-central Tanzania showed results similar to the laboratory and semi-field studies and suggested that larval habitat volatiles also attract other disease vectors (Chapter 6). This study showed that infochemicals from selected sites can induce an oviposition response in mosquitoes under natural conditions, and that soil from a breeding site and nonane both induced oviposition in An. gambiae and Cx. quinquefasciatus with a synergistic effect caused by nonane + soil in An. gambiae. Therefore, the study also revealed that the oviposition-site-selection process of gravid mosquitoes is mediated by a combination of infochemicals derived from natural breeding sites. This knowledge can be exploited to develop a control strategy based on the push-pull or attract-and-kill concept. Furthermore, the knowledge can be utilized to develop other surveillance tools for mosquito vectors of human and animal diseases.

In order to explore the potential of using a push-pull system for mosquito control, the attractants were combined with environmentally friendly biolarvicides in order to achieve more effective control of their aquatic larval offspring (Chapter 7). As an attractant for gravid mosquitoes, nonane, which originated from An. gambiae larvae, was selected. As toxicant for mosquito larvae, Bacillus thuringiensis var israeliensis (Bti) and Temephos were selected. Mosquitoes were successfully attracted to lay eggs in water-filled clay pots treated with attractants and combined with larvicides. Therefore, a lure-and-kill system was successfully implemented; mosquitoes were attracted to lay eggs in intervention pots containing toxicants, consequently, all emerged larvae died before reaching the pupal stage.

The results presented in this thesis create new opportunities for applied research, especially on the development of novel vector-borne disease control strategies using infochemicals (Chapter 8). The volatiles produced by the larval habitats play an important role in the site-selection behaviour of An. gambiae and Cx. quinquefasciatus mosquitoes. The attractiveness of the synthetic chemicals identified in this study may be improved by making blends of attractants or in combination with other infochemicals identified elsewhere. Identified compounds can be used in push-pull and/or lure-and-kill strategies for the manipulation of mosquitoes; thereby reducing the abundance of mosquitoes, the human-biting frequency, and the intensity of pathogen transmission. Research on the role of oviposition-based infochemicals in the site-seeking behaviour of mosquitoes may lead to a better understanding of vector behaviour and contribute to the fight against vector-borne diseases.

The main conclusions from this thesis can be summarized as follows:

Volatiles produced by larval habitats play an important role in the mediation of oviposition behaviour of An. gambiae and in the dispersal of mosquito species. The analysis of the headspace from larval microcosms led to the identification of 16 compounds. Out of these, four (DMDS, DMTS, nonane and 2,4-pentanedione) had behavioural effects on An. gambiae. Nonane acts as an oviposition cue for Cx. quinquefasciatus as well. Mosquitoes are governed by multiple cues in the selection of suitable breeding sites, thus further research on the role of combined infochemicals to optimize selection of breeding sites should be initiated. These infochemicals can be used in push-pull or attract–and-kill strategies for surveillance and control of malaria mosquito larvae. Thus, odour-based technologies can improve the surveillance, sampling and control strategies for disease vectors. In this way the burden resulting from mosquito-borne diseases such as morbidity, mortality and economic losses will be decreased.