Understanding the long-lasting attraction of malaria mosquitoes to odor baits

Abstract

The use of odor baits for surveillance and control of malaria mosquitoes requires robust dispensing tools. In this study, the residual activity of a synthetic mosquito attractant blend dispensed from nylon or low density polyethylene (LDPE) sachets was evaluated at weekly intervals for one year without re-impregnation. The potential role of bacteria in modulating the attraction of mosquitoes to odor-treated nylon that had been used repeatedly over the one year study period, without re-impregnation, was also investigated. Significantly higher proportions of female Anopheles gambiae sensu stricto mosquitoes were consistently attracted to treated nylon strips than the other treatments, up to one year post-treatment. Additional volatile organic compounds and various bacterial populations were found on the treated nylon strips after one year of repeated use. The most abundant bacteria were Bacillus thuringiensis and Acinetobacter baumannii. Autoclaving of treated nylon strips prior to exposure had no effect on trap collections of laboratory-reared female An. Gambiae (P = 0.17) or wild female An. Gambiae sensu lato (P = 0.26) and Mansonia spp. (P = 0.17) mosquitoes. Trap catches of wild female An. Funestus (P < 0.001) and other anophelines (P < 0.007) were higher when treated strips had been autoclaved prior to deployment as opposed to when the treated nylon strips were not autoclaved. By contrast, wild female Culex mosquitoes were more strongly attracted to non-autoclaved compared to autoclaved treated nylon strips (P < 0.042). This study demonstrates the feasibility of using odor baits for sampling and surveillance of malaria as well as other mosquito vectors over prolonged periods of time. Preliminary evidence points towards the potential role of bacteria in sustaining prolonged use of nylon material for dispensing synthetic attractant odorants for host-seeking malaria and other mosquito vectors but further investigations are required.

Fig 1. Proportions of female An. gambiae mosquitoes caught weekly for one year under semi-field conditions in traps containing IB1-treated nylon strips (○), LDPE sachets filled with the IB1 odour blend (●), control nylon strips (□) and control LDPE sachets (■).

The treatments were used repeatedly for 52 nights without refreshing them. The solid, dotted, dashed and dashed-with-square lines represent the baseline-category logit model fit showing trends of proportions of mosquitoes attracted over time.

Fig 2. Bacterial culture plates showing results of streaking with IB1-treated nylon strips used repeatedly for collecting female An. gambiae mosquitoes (one night per week) for one year (panel A), strips of a nylon sock worn for 12 h by a human volunteer (panel B) and control (untreated) nylon strips (panel C).

Fig 3. Percent of female Anopheles gambiae mosquitoes choosing either treatment A or B in a dual choice olfactometer assay.

Blends of volatile organic compounds were released from nutrient agar blocks inoculated with B. thuringiensis and A. baumannii isolates or nutrient agar blocks only. Each treatment was tested four times over two days. The total number of mosquitoes released was 120. The total number of mosquitoes trapped per choice test (n) is shown. Error bars represent standard errors of the mean mosquito percentage. Asterisks indicate significant differences between treatment A and B; ** P < 0.01; *** P < 0.001 (χ² test).

Fig 4. Percent of female Anopheles gambiae mosquitoes choosing either treatment A or B in a semi-field dual choice assay.

The effect of autoclaving on attraction of female Anopheles gambiae mosquitoes to untreated (―) and odour baited (blend IB1) (+) nylon strips was assessed. Nylon strips were either non-autoclaved (―) or autoclaved (+). The total number of mosquitoes released (N) for each dual choice test, replicated during four nights, was 800. The total number of mosquitoes trapped per choice test (n) is shown. Error bars represent standard errors of the mean. Bars followed by * denote significant differences (P < 0.05) in mosquito catches between treatments.