Wolbachia incompatible insect technique program optimization over large spatial scales using a process-based model of mosquito metapopulation dynamics

Abstract

Background: Wolbachia incompatible insect technique (IIT) programs have been shown in field trials to be highly effective in suppressing populations of mosquitoes that carry diseases such as dengue, chikungunya, and Zika. However, the frequent and repeated release of Wolbachia-infected male mosquitoes makes such programs resource-intensive. While the need for optimization is recognized, potential strategies to optimize releases and reduce resource utilization have not been fully explored.

Results: We developed a process-based model to study the spatio-temporal metapopulation dynamics of mosquitoes in a Wolbachia IIT program, which explicitly incorporates climatic influence in mosquito life-history traits. We then used the model to simulate various scale-down and redistribution strategies to optimize the existing program in Singapore. Specifically, the model was used to study the trade-offs between the intervention efficacy outcomes and resource requirements of various release program strategies, such as the total number of release events and the number of mosquitoes released. We found that scaling down releases in existing sites from twice a week to only once a week yielded small changes in suppression efficacy (from 87 to 80%), while requiring 44% fewer mosquitoes and release events. Additionally, redistributing mosquitoes from already suppressed areas and releasing them in new areas once a week led to a greater total suppressive efficacy (83% compared to 61%) while also yielding a 16% and 14% reduction in the number of mosquitoes and release events required, respectively.

Conclusions: Both scale-down and redistribution strategies can be implemented to significantly reduce program resource requirements without compromising the suppressive efficacy of IIT. These findings will inform planners on ways to optimize existing and future IIT programs, potentially allowing for the wider adoption of this method for mosquito-borne disease control.

Keywords: Aedes aegypti; Wolbachia; Compartmental model; Incompatible insect technique; Mosquitoes; Neglected tropical diseases; Simulation.

Fig. 1

Model-based simulation (a) and the empirically estimated (b) intervention efficacies of the early release program in Singapore. The x-axis represents time since the intervention began. The model-based intervention efficacy was defined as the difference in the adult ovipositing female abundance between the intervention and baseline (no intervention) simulations. Each thin line represents the intervention efficacy for a single hexagon, whereas each thick line represents the mean intervention efficacy for an entire hexagon class. Core, buffer, and adjacent non-release hexagons are colored in blue, orange, and gray, respectively. The mean intervention efficacy line for all release hexagons (core and buffer) is shown in red

Fig. 2

Intervention efficacy outcomes of various scale-down strategies. Plots a, b, and c show the mean intervention efficacy in all release, core-only, and buffer-only hexagons respectively for all scale-down strategies. The remaining plots d to i show the detailed intervention efficacies in all hexagons for Strategies 1 to 6, respectively

Fig. 3

Intervention efficacy outcomes of various redistribution strategies. The first row of plots shows the mean intervention efficacy in both current and new hexagons (a), current hexagons (b), and new hexagons only (c). The remaining plots d, e, and f show the intervention efficacies in all hexagons for Strategies 1, 7, and 8, respectively

Fig. 4

Model diagram for Wolbachia-positive mosquitoes across various life cycle stages

Fig. 5

Model diagram for Wolbachia-negative mosquitoes across various life cycle stages

Fig. 6

The spatial distribution of mosquitoes was represented by a grid of vertically aligned hexagons. Mosquitoes mix freely within each hexagon (left). Some mosquitoes move between neighboring hexagons at each timestep (right)

Fig. 7

Release hexagons modeled after the early release program in Singapore. Core, buffer, and adjacent non-release hexagons are colored in blue, orange, and gray, respectively

Fig. 8

Release hexagons for the current release program in Singapore, core and buffer release hexagons colored in blue and orange, respectively (a). Release hexagons for the proposed expanded release program in Singapore, current and new release hexagons colored in blue and purple, respectively (b)