Design and Testing of Novel Lethal Ovitrap to Reduce Populations of Aedes Mosquitoes: Community-Based Participatory Research between Industry, Academia and Communities in Peru and Thailand

Abstract

Background: Dengue virus (and Chikungunya and Zika viruses) is transmitted by Aedes aegypti and Aedes albopictus mosquitoes and causes considerable human morbidity and mortality. As there is currently no vaccine or chemoprophylaxis to protect people from dengue virus infection, vector control is the only viable option for disease prevention. The purpose of this paper is to illustrate the design and placement process for an attractive lethal ovitrap to reduce vector populations and to describe lessons learned in the development of the trap.

Methods: This study was conducted in 2010 in Iquitos, Peru and Lopburi Province, Thailand and used an iterative community-based participatory approach to adjust design specifications of the trap, based on community members' perceptions and feedback, entomological findings in the lab, and design and research team observations. Multiple focus group discussions (FGD) were held over a 6 month period, stratified by age, sex and motherhood status, to inform the design process. Trap testing transitioned from the lab to within households.

Results: Through an iterative process of working with specifications from the research team, findings from the laboratory testing, and feedback from FGD, the design team narrowed trap design options from 22 to 6. Comments from the FGD centered on safety for children and pets interacting with traps, durability, maintenance issues, and aesthetics. Testing in the laboratory involved releasing groups of 50 gravid Ae. aegypti in walk-in rooms and assessing what percentage were caught in traps of different colors, with different trap cover sizes, and placed under lighter or darker locations. Two final trap models were mocked up and tested in homes for a week; one model was the top choice in both Iquitos and Lopburi.

Discussion: The community-based participatory process was essential for the development of novel traps that provided effective vector control, but also met the needs and concerns of community members.

Fig 1. Participatory and interactive model depicting design approach.

Model depicting the iterative participatory approach throughout the six phases of the trap design process.

Fig 2. Depiction of second round of trap prototypes.

Six trap prototypes presented and discussed in nine focus group discussions in Peru and Thailand.

Fig 3. Depiction of third round of trap prototypes.

Four trap prototypes (J, H, P, M, respectively) presented as images and mock-ups in six focus groups in Peru and in Thailand for discussion and ranking.

Fig 4. Evaluation of cans with stripe patterns compared to black cans.

Stripe patterns on black cans evaluated in walk-in cage, sticky panel bioassays. Cans were evaluated at the same time.

Fig 5. Depiction of evaluation of various trap prototypes.

Top panel: Walk-in bioassay room showing placement of red cover and black cover cans. Bottom panel: Black cans fitted with small, medium and large sized covers (left to right).

Fig 6. Depiction of evaluation of efficacy of two trap models.

H (oval) and J (lantern) traps compared to black tin can with cover in walk-in bioassay room.

Fig 7. Depiction of Model P prototype.

Model P prototype trap showing Duranet® netting.

Fig 8. Evaluation of red trap covers of different sizes.

Mean percentage (± SD) of gravid NOLA strain Aedes aegypti caught on sticky panels in black cans fitted with different sized red covers. Each trap with a red cover was tested against a black one gallon tin can fitted with a black cover. Traps were placed 1 m apart in the center of a walk-in cage (n = 4).

Fig 9. Evaluation of trap prototype based on room placement in the walk-in room.

Mean percentage (± SD) of gravid NOLA strain Aedes aegypti caught on sticky panels in red striped cans with black tops placed 1 m apart in the center or in corners of a walk-in room (n = 4). Percentages represent females that were trapped in the experimental trap relative to the total number of females trapped in all four traps used in each assay, i.e., the sum of all percentages for each experimental design equals 100%.

Fig 10. Evaluation of trap prototype based on exposure to different light levels in walk-in bioassay cage.

Mean percentage (± SD) of gravid NOLA strain Ae. aegypti caught on sticky screens in red and black cans placed in diagonal corners of walk-in bioassay cages. One corner was dimly lit while the opposite corner had normal light levels.

Fig 11. Evaluation of different prototypes with and without funnels, and with different volumes of well water.

Mean percentage (± SD) of gravid NOLA strain Ae. aegypti caught on sticky screens in HLB “H” and “J” traps in walk-in cage bioassays (n = 4). Some assays were the two-choice and included a one gallon black can (n = 4).

Fig 12. Evaluation of gravid Iquitos Ae. aegypti mortality based on bacterial bead lyophilization date.

Mean percentage (± SD) mortality of gravid Iquitos strain Ae. aegypti (n = 50) exposed for 24-h to HLB second “H” trap prototype in walk-in bioassay cages. Each trap contained 500 mL of Raleigh tap water and the contents of one A&K pouch (80 mg spinosad and 100 mg bacterial beads). Dates correspond to the dates that bacterial beads were lyophilized.

Fig 13. Design of final prototype.