The role of grass volatiles on oviposition site selection by Anopheles arabiensis and Anopheles coluzzii

Abstract

Background: The reproductive success and population dynamics, of Anopheles malaria mosquitoes is strongly influenced by the oviposition site selection of gravid females. Mosquitoes select oviposition sites at different spatial scales, starting with selecting a habitat in which to search. This study utilizes the association of larval abundance in the field with natural breeding habitats, dominated by various types of wild grasses, as a proxy for oviposition site selection by gravid mosquitoes. Moreover, the role of olfactory cues emanating from these habitats in the attraction and oviposition stimulation of females was analysed.

Methods: The density of Anopheles larvae in breeding sites associated with Echinochloa pyramidalis, Echinochloa stagnina, Typha latifolia and Cyperus papyrus, was sampled and the larvae identified to species level. Headspace volatile extracts of the grasses were collected and used to assess behavioural attraction and oviposition stimulation of gravid Anopheles arabiensis and Anopheles coluzzii mosquitoes in wind tunnel and two-choice oviposition assays, respectively. The ability of the mosquitoes to differentiate among the grass volatile extracts was tested in multi-choice tent assays.

Results: Anopheles arabiensis larvae were the most abundant species found in the various grass-associated habitats. The larval densities described a hierarchical distribution, with Poaceae (Echinochloa pyramidalis and Echinochloa stagnina)-associated habitat sites demonstrating higher densities than that of Typha-associated sites, and where larvae were absent from Cyperus-associated sites. This hierarchy was maintained by gravid An. arabiensis and An. coluzzii mosquitoes in attraction, oviposition and multi-choice assays to grass volatile extracts.

Conclusions: The demonstrated hierarchical preference of gravid An. coluzzii and An. arabiensis for grass volatiles indicates that vegetation cues associated with larval habitats are instrumental in the oviposition site choice of the malaria mosquitoes. Identifying volatile cues from grasses that modulate gravid malaria mosquito behaviours has distinct potential for the development of tools to be used in future monitoring and control methods.

Keywords: Anopheles arabiensis; Anopheles coluzzii; Attraction; Habitat; Olfaction; Oviposition; Selection.

Fig. 1

Anopheles larval density in natural breeding habitats dominated by four grass species, Echinochloa pyramidalis, Echinochloa stagnina, Typha latifolia and Cyperus papyrus, assayed in early (top) and late (bottom) September. The mean larval densities with different letter designations are significantly different from one another (univariate general linear model with a Tukey’s post hoc analysis; P < 0.005). Error bars represent the standard error of the mean

Fig. 2

Attraction of gravid Anopheles arabiensis and Anopheles coluzzii to grass volatile headspace extracts, from Echinochloa pyramidalis (a, e) Echinochloa stagnina (b, f), Typha latifolia (c, g) and Cyperus papyrus (d, h), respectively. Release rate of the volatile headspace extracts is given in minute equivalents. The solvent control was hexane (HEX). The mean percent attraction values with the same letters indicate no significant difference from one another (nominal logistic fit model, χ² and P < 0.05 from the likelihood ratio test, significant differences were determined by odds ratio pairwise comparisons). Error bars represent the standard error of the mean

Fig. 3

Oviposition response of gravid Anopheles arabiensis and Anopheles coluzzii to grass volatile headspace extracts, from Echinochloa pyramidalis (a, e) Echinochloa stagnina (b, f), Typha latifolia (c, g) and Cyperus papyrus (d, h), respectively. Release rate of the volatile headspace extracts is given in minute equivalents. The solvent control was hexane (HEX). Mean oviposition preferences with the same letters indicate no significant difference from one another (nominal logistic fit model, χ² and P < 0.05 from the likelihood ratio test, significant differences were determined by odds ratio pairwise comparisons). Error bars represent the standard error of the mean

Fig. 4

Oviposition preference hierarchy of Anopheles arabiensis (top) and Anopheles coluzzii (bottom) in a multi-choice arena, to water conditioned with volatile headspace extracts of the four grass species, Echinochloa pyramidalis, Echinochloa stagnina, Typha latifolia and Cyperus papyrus. Mean numbers of eggs with the same letters indicate no significant difference from one another (nominal logistic fit model, χ² and P < 0.05 from the likelihood ratio test, significant differences were determined by odds ratio pairwise comparisons). Error bars represent the standard error of the mean