Skip to main content

The ecology and population dynamics of Wolbachia infections in Aedes aegypti and the development of new strategies for mosquito-borne disease control

Final Report Summary - WOLBACHIA_MOD (The ecology and population dynamics of Wolbachia infections in Aedes aegypti and the development of new strategies for mosquito-borne disease control)

Final Report: The ecology and population dynamics of Wolbachia infections in Aedes aegypti and the development of new strategies for mosquito-borne disease control

Project website:
Project blog:

Contact: Penelope Hancock. Email:

Main Aims:
The two major Aims stated for our project are:
Aim 1: To experimentally examine the effect of Wolbachia infection on the competitive ability of A. aegypti larvae across a realistic range of habitat conditions.
Aim 2: To develop data-driven mathematical models to predict strategies for the release of Wolbachia into A. aegypti populations that are robust to endogenous and exogenous environmental variation.

Aim 1:
To meet the first Aim, we designed experimental studies of the dynamics of wMel Wolbachia in populations of A. aegypti housed in field-cages. We studied three experimental field-cage populations for a period of about 6 months each. We used these observations to estimate the variation in important fitness components of infected and uninfected mosquitoes over time, including larval survival, larval development times, and adult female fecundity.
We found that mosquito fitness components varied strongly over time in association with changing larval density in the field-cage populations. Fitness components did not differ significantly between Wolbachia-infected and uninfected mosquitoes, indicating that Wolbachia infection did not affect mosquito fitness. However, density-dependent variation in mosquito fitness components strongly affected the rate of Wolbachia spread. Specifically, the intense larval density-dependent competition occurring in the experimental populations caused marked extension of larval development times, which slowed the rate of Wolbachia spread. Experimental populations that received the same larval food supply regime showed similar patterns of density-dependent variation in fitness components, which supports the reproducibility of our results.

Aim 2:
We incorporated the empirical relationships describing the form of density-dependent variation in mosquito fitness components into mathematical models to predict the dynamics of wMel Wolbachia field release strategies. We used two main modeling approaches: the first approach ignores spatial structure and assumes the mosquito population is well-mixed, and the second approach adopts a spatially explicit representation of the mosquito population and its habitat.
We used our models to characterize the dynamics of wMel field release strategies across varying intensities of density-dependent competition that may be experienced in field mosquito populations. We identified three main ways in which varying competition intensity in field populations can dramatically affect the dynamics of the field releases. First, when the field population experiences intense competition, released mosquitoes have a large fitness advantage due to the relatively high fecundity of adult females. This means that fewer released mosquitoes are required to achieve a target Wolbachia infection frequency at the end of a release program. Second, the development rate of mosquito larvae is also strongly density-dependent which affects the speed of Wolbachia invasion. Third, transient Wolbachia dynamics associated with field releases are influenced by density-dependent mosquito population growth rates. These effects have greater impacts on dynamics when released mosquitoes experience fitness disadvantages that can arise from being maladapted to the field environment.
Extending our model to consider spatial structure in the mosquito population (by subdividing the population into houses across a 5km2 area), we showed that our predictions of the rate of spatial spread of wMel agree well with observations from field releases of wMel conducted in northeast Australia. We then analysed release strategies involving different spatial distributions of released mosquitoes, and found that releasing mosquitoes into high quality habitats achieved the fastest rate of Wolbachia invasion across the area. Releasing mosquitoes into spatial clumps (blocks of houses) was much less efficient in achieving area-wide Wolbachia invasion.

We have developed novel approaches to modeling Wolbachia invasion dynamics in Aedes aegypti mosquito populations informed by new experimental data describing density-dependent variation in mosquito demographic traits. Our models accurately explain observed Wolbachia dynamics, and provide a new perspective on the dynamics of Wolbachia field release strategies for arbovirus control. Specifically, we show how demographic variation in mosquito populations due to density-dependent processes can greatly affect invasion outcomes following field releases. We have produced publically available computational tools for use in predicting Wolbachia field release dynamics across varying environmental conditions.

Socio-economic impacts
Field releases of Wolbachia bacteria to assist in the control of arboviruses transmitted by Aedes aegypti mosquitoes are currently being implemented in several countries across multiple continents, including South America, southeast Asia and Australia. International collaborative partnerships such as Eliminate Dengue ( are coordinating these field trials. Our models will be available to Eliminate Dengue, as well as other researchers and organizations involved in conducting Wolbachia field releases. To increase relevance to public health managers, we consider spatial dimensions that represent current strategic goals of scaling up field release strategies to combat arbovirus transmission across large urban areas.