Mosquito host preferences affect their response to synthetic and natural odour blends

Abstract

Background: The anthropophilic malaria mosquito Anopheles gambiae sensu stricto (hereafter termed Anopheles gambiae) primarily takes blood meals from humans, whereas its close sibling Anopheles arabiensis is more opportunistic. Previous studies have identified several compounds that play a critical role in the odour-mediated behaviour of An. gambiae. This study determined the effect of natural and synthetic odour blends on mosquitoes with different host preferences to better understand the host-seeking behaviour of mosquitoes and the potential of synthetic odour blends for standardized monitoring.

Methods: Odour blends were initially tested for their attractiveness to An. gambiae and An. arabiensis in a semi-field system with MM-X traps baited with natural and synthetic odours. Natural host odours were collected from humans, cows and chickens. The synthetic odour blends consisted of three or five previously identified compounds released with carbon dioxide. These studies were continued under natural conditions where odour blends were tested outdoors to determine their effect on species with different host preferences.

Results: In the semi-field experiments, human odour attracted significantly higher numbers of both mosquito species. However, An. arabiensis was also attracted to cow and chicken odours, which confirms its opportunistic behaviour. A five-component synthetic blend was highly attractive to both mosquito species. In the field, the synthetic odour blend caught significantly more An. funestus than traps baited with human odour, while no difference was found for An. arabiensis. Catches of An. arabiensis and Culex spp. contained large numbers of blood-fed mosquitoes, mostly from cows, which indicates that these mosquitoes had fed outdoors.

Conclusions: Different odour baits elicit varying responses among mosquito species. Synthetic odour blends are highly effective for trapping mosquitoes; however, not all mosquitoes respond equally to the same odour blend. Combining fermenting molasses with synthetic blends in a trap represents the most effective tool to catch blood-fed mosquitoes outside houses, which is essential for understanding outdoor malaria transmission.

Figure 1

Screenhouse mosquito catches in dual-choice test with different odour baits. Estimated mean proportion (GLM) of mosquitoes caught in a screenhouse using MM-X traps with CO₂ tested versus an empty trap (A). Cow (B) chicken C) and human (D) emanations were tested in combination with CO₂ versus a trap with CO₂ alone. Error bars represent the standard error of the mean; ***: χ²-test P < 0.001, NS: χ²-test P > 0.05. Numbers in the bars indicate number of mosquitoes caught.

Figure 2

Competition experiment in a screenhouse with traps baited with natural odours from different host species. Estimated mean proportion (GLM) of mosquitoes caught in a screenhouse using MM-X traps with CO₂ only (control), or CO₂ and cow, chicken or human odours. Error bars represent the standard error of the mean. Numbers in the bars indicate number of mosquitoes caught. For each mosquito species: different letters indicate significant differences between treatments for each mosquito species (P < 0.05, GLM).

Figure 3

Screenhouse mosquito catches in traps baited with synthetic blends. Estimated mean proportion (GLM) of mosquitoes caught in a screenhouse using MM-X traps without (strips only) and with CO₂ (control) or with CO₂ plus synthetic blends. Error bars represent the standard error of the mean. Numbers in the bars indicate number of mosquitoes caught. For each mosquito species: different letters indicate significant differences between treatments (P < 0.05, GLM).

Figure 4

Mosquito catches in traps baited with natural and synthetic odours in a field set-up. Estimated mean proportion (GLM) of wild unfed or blood-fed mosquitoes caught outdoors. A) An. arabiensis, B) An. funestus s.l. caught outdoors using MM-X traps with CO₂ or CO₂ and treatments. Numbers in the bars indicate number of mosquitoes caught during 25 experimental nights. For each mosquito species: different letters indicate significant differences between treatments (P < 0.05, GLM).