Discovery of an oviposition attractant for gravid malaria vectors of the Anopheles gambiae species complex

Abstract

Background: New strategies are needed to manage malaria vector populations that resist insecticides and bite outdoors. This study describes a breakthrough in developing 'attract and kill' strategies targeting gravid females by identifying and evaluating an oviposition attractant for Anopheles gambiae s.l.

Methods: Previously, the authors found that gravid An. gambiae s.s. females were two times more likely to lay eggs in lake water infused for six days with soil from a natural oviposition site in western Kenya compared to lake water alone or to the same but autoclaved infusion. Here, the volatile chemicals released from these substrates were analysed with a gas-chromatograph coupled to a mass-spectrometer (GC-MS). Furthermore, the behavioural responses of gravid females to one of the compounds identified were evaluated in dual choice egg-count bioassays, in dual-choice semi-field experiments with odour-baited traps and in field bioassays.

Results: One of the soil infusion volatiles was readily identified as the sesquiterpene alcohol cedrol. Its widespread presence in natural aquatic habitats in the study area was confirmed by analysing the chemical headspace of 116 water samples collected from different aquatic sites in the field and was therefore selected for evaluation in oviposition bioassays. Twice as many gravid females were attracted to cedrol-treated water than to water alone in two choice cage bioassays (odds ratio (OR) 1.84; 95% confidence interval (CI) 1.16-2.91) and in experiments conducted in large-screened cages with free-flying mosquitoes (OR 1.92; 95% CI 1.63-2.27). When tested in the field, wild malaria vector females were three times more likely to be collected in the traps baited with cedrol than in the traps containing water alone (OR 3.3; 95% CI 1.4-7.9).

Conclusion: Cedrol is the first compound confirmed as an oviposition attractant for gravid An. gambiae s.l. This finding paves the way for developing new 'attract and kill strategies' for malaria vector control.

Figure 1

Experimental set-up. (A) Cage bioassays with individual gravid females under ambient conditions in makeshift huts; (B) Modified BG sentinel traps in a semi-field system; (C) Field set-up of square of electrocuting nets (up) and OviART gravid trap (down).

Figure 2

Example chromatograms from round five of volatile collections. One chromatogram of each sample type (unmodified soil infusion, autoclaved soil infusion and lake water) and empty bottle control. All compounds included in the multivariate analysis are marked by the corresponding ID number. Kovarts retention index (RI) and mass spectral data for each compound can be found in Additional file 1.

Figure 3

Biplot of the GC-MS data from lake water, unmodified and autoclaved soil infusions. The three sample types form distinct groups, mainly separated by the second principal component. Four compound IDs (51, 263, 276 and 283) group closely with the unmodified soil samples. Data from seven rounds of each sample type were centred and standardized by the volatile compounds before being subjected to principal component analysis with supplementary variables. The supplementary variables were the three sample types indicated with WATER (lake water), AUTO (autoclaved soil infusion) and SOIL (unmodified soil infusions). Each sample is indicated with a letter; W, A or S for lake water, autoclaved soil infusion and unmodified soil infusion respectively. The number following the letter indicates the round; volatiles were collected in parallel from samples with the same number.

Figure 4

Mean per cent of gravid Anopheles gambiae responding to control and test treatments in choice experiments. (A) Cage bioassays with soil infusions of increasing incubation time and comparison of autoclaved versus unmodified infusion. The data from Herrera-Varela and others [12] have been re-analysed for this figure to show the per cent of females responding. These data present the background for the current study. Headspace collections for identification of volatile chemicals were implemented for autoclaved and unmodified six-day old soil infusions in parallel to these behavioural assays. (B) Cage bioassays with cedrol-treated lake water in increasing concentrations. (C) Semi-field evaluation of response off free-flying gravid females to cedrol-baited traps.

Figure 5

Estimated mean number of female mosquitoes per trap night (all trap types pooled) collected during the field trial. Error bars represent 95% confidence intervals.