Skip to main content
  • Loading metrics

Application of wMelPop Wolbachia Strain to Crash Local Populations of Aedes aegypti

  • Scott A. Ritchie ,

    Affiliations College of Public Health, Medical and Veterinary Sciences, Australian Institute of Tropical Health and Medicine, James Cook University, Cairns, Queensland, Australia, Australian Institute of Tropical Health and Medicine, James Cook University, Cairns, Queensland, Australia

  • Michael Townsend,

    Affiliations College of Public Health, Medical and Veterinary Sciences, Australian Institute of Tropical Health and Medicine, James Cook University, Cairns, Queensland, Australia, Australian Institute of Tropical Health and Medicine, James Cook University, Cairns, Queensland, Australia

  • Chris J. Paton,

    Affiliations College of Public Health, Medical and Veterinary Sciences, Australian Institute of Tropical Health and Medicine, James Cook University, Cairns, Queensland, Australia, Australian Institute of Tropical Health and Medicine, James Cook University, Cairns, Queensland, Australia

  • Ashley G. Callahan,

    Affiliation Pest and Disease Vector Group, Department of Genetics, Bio21 Institute, The University of Melbourne, Melbourne, Victoria, Australia

  • Ary A. Hoffmann

    Affiliation Pest and Disease Vector Group, Department of Genetics, Bio21 Institute, The University of Melbourne, Melbourne, Victoria, Australia


The endosymbiotic bacteria Wolbachia pipientis (wMel strain) has been successfully established in several populations of Aedes aegypti, the primary dengue vector. The virulent Wolbachia strain wMelPop is known to cause several pathological impacts (increased egg mortality, life shortening, etc.) reducing overall fitness in the mosquito Ae. aegypti. Increased egg mortality could substantially reduce egg banks in areas with a lengthy monsoonal dry season, and be employed to eliminate local populations. We tested this application under semi-field cage conditions. First, we determined that wMelPop infection significantly reduced the survival of desiccation-resistant eggs of the dengue vector Ae. aegypti, with shade and temperature having a significant impact; nearly all wMelPop-infected eggs failed to hatch after 6 and 10 weeks in summer and winter conditions, respectively. In laboratory selection experiments we found that egg desiccation resistance can be increased by selection, and that this effect of wMelPop infection is due to the nuclear background of the host rather than Wolbachia. We then conducted an invasion of wMelPop within a semi-field cage using sustained weekly releases of wMelPop infected mosquitoes, with fixation achieved after 9 weeks. The egg populations wMelPop infected and an uninfected control were then subjected to a simulated prolonged monsoonal dry season (2.5 months) before flooding to induce hatching. The wMelPop infected eggs suffered significantly greater mortality than the controls, with only 0.67% and 4.35% of respective infected and uninfected eggs held in 99% shade hatching after 80 days. These studies suggest that wMelPop could be used to locally eliminate populations of Ae. aegypti that are exposed to prolonged dry conditions, particularly if combined with vector control.

Author Summary

Dengue is a leading cause of morbidity in the tropics. As a commercial vaccine is not available, control or modification of the mosquito vectors is employed to prevent transmission. Strains of the endosymbiotic bacteria Wolbachia affect the survival and ability of the dengue vector Aedes aegypti to transmit dengue. The Wolbachia strain wMelPop over-replicates within Ae. aegypti, inducing strong dengue virus blocking and early mortality of both egg and adult mosquitoes. We investigated whether this life-shortening Wolbachia strain can be used to eliminate local populations of Ae. aegypti in a semi-field cage. Our results indicate that Ae. aegypti eggs infected with wMelPop died at a significantly higher rate than uninfected eggs, and were nearly eliminated during a simulated dry season of 2–3 months. This suggests that that releases of wMelPop could facilitate control and elimination of Ae. aegypti if used in concert with vector control.


Dengue is the leading cause of arboviral disease in humans. An estimated 390 million infections and 90 million clinical cases occur annually [1]. There is no commercially available vaccine, so vector control and mosquito avoidance are the only methods to limit transmission. Traditional control of the mosquito vector Aedes aegypti is increasingly hampered by physiological resistance against insecticides, especially synthetic pyrethroids [25]. Furthermore, the logistics of controlling a vector that exploits artificial containers that are often cryptic, subterranean and difficult to access (eg., elevated sites such as rain water tanks) makes effective control difficult to achieve [68]. The adult Ae. aegypti is endophilic, and is often found inside dark, quiet areas within premises [9]. These areas are not effectively targeted by insecticidal sprays applied by air or ground [9,10]. Collectively, vector control efforts to control dengue are generally not successful, and dengue continues to expand in breadth and reach globally.

New vector population manipulation methods have been developed to overcome the issues of traditional vector control. The release of insects infected with a dominant lethal gene (RIDL) employs the release of male Ae. aegypti that induce sterility by passing a lethal gene to wild females they mate with [11]. Serial releases of these genetically modified male mosquitoes thus lead to collapse of the resident wild population [11,12]. The vector capacity of resident populations of Ae. aegypti can also be reduced to lessen dengue transmission. Strains of the endosymbiotic bacteria Wolbachia were transinfected into Ae. aegypti eggs, and are then passed on maternally [13]. Male Ae. aegypti infected with Wolbachia induce embryo death when mating with uninfected females [13,14]. This creates a powerful drive mechanism that allows Wolbachia to spread naturally within a population of uninfected Ae. aegypti, and to persist once fixed [1517]. The presence of Wolbachia infection also interrupts dengue (and other arboviruses) replication in Ae. aegypti, interfering with transmission [1820].

Two virus blocking strains of Wolbachia (wMel and wMelPop) have been established in Ae. aegypti. To date, populations of wMel-infected Ae. aegypti have been released and established in seven localities around Cairns, Australia. The more virulent wMelPop over-replicates inside the mosquito, inducing a variety of physiological manifestations. These include early death (life-shortening) [13], egg mortality [21,22], reduced blood feeding [23], delayed larval development [24] and reduced overall fitness [25]. The high density of wMelPop probably contributes to its almost complete blocking of dengue replication and transmission in Ae. aegypti [18], but efforts to introduce this infection into wild populations have proved challenging [25], although it has been accomplished in semi-field cages [19]. One implication of high mortality, especially reduced egg longevity, is that wMelPop could lead to reduced population size of Ae. aegypti [26]. In areas with a pronounced monsoonal dry season, mosquitoes survive for up to several months as desiccation resistant eggs [27]. If wMelPop can be established in these populations during the wet season, populations might naturally die off during the dry season.

In this paper we measure the relative survival of wMelPop infected Ae. aegypti eggs under natural conditions. Because this strategy could be thwarted if there is evolution in wMelPop to counter any deleterious fitness effect, we also explore strain variation involved in egg survival. Finally, we simulate a wMelPop intervention from invasion to subsequent death of the eggs in a semi-field cage to test the concept that wMelPop could be used to crash and locally eliminate Ae. aegypti egg banks under a prolonged monsoonal dry season.


All mosquito colonies were maintained using the standard laboratory rearing protocols [22] with minor modifications. For the egg survival experiment we used eggs from a wMelPop colony derived from the original microinjected wMelPop—CLA line which had been backcrossed to wild Cairns material for 7 generations and maintained in the laboratory with several hundred adults turned over each generation [28]. This strain was compared to uninfected material derived directly from adults or their offspring emerging from field collections of Ae. aegypti eggs. The Ae. aegypti mosquito strain used for the elimination trial and to create the desiccation selected line was a field-sourced population of the wMelPop Wolbachia infected line (AOMB) [24]. This colony was initiated from 18 infected females collected from Machans Beach 0.6 months after releases of wMelPop had been terminated there. The infection in this colony had therefore been exposed to field conditions during breeding cycles. It was subsequently maintained in the laboratory at >200 adults. Experiments with this colony [24] indicated that the deleterious host fitness effects that wMelPop is known to have on longevity, fecundity and egg viability were also detectable in this colony. Prior to the start of experiments, AOMB was backcrossed for three generations with adults emerging from wildtype eggs from Cairns, QLD.

Egg survival

A protective awning with three different shade regimes was built to house mosquito eggs in the survival study. The awning (17 m x 3.3 m) was built between 2 semi-field cages [29] and divided into three sections that provided 30%, 70% and 99% shade (Fig 1). The awning roof consisted of waterproof translucent vinyl to protect eggs from rainfall. Two 5.6 m sections of the vinyl had an interior piece of 40% and 95% shade cloth to provide additional shade to create the 70% and 99% shade; the section with vinyl alone provided 30% shade. Temperatures and relative humidity within each section were monitored using data loggers (Hygrochron iButton DS1923, Dallas Semiconductors) set atop and below the table, supplemented with data from the Bureau of Meteorology (BOM) site 15 km SE of the site when loggers failed. Preliminary trials indicated that ambient temperatures (under table) in the 70% and 99% shade (S1 Fig) were nearly identical to screen height temperatures at the Cairns Airport BOM site. However, loggers on the table that were exposed to sunlight (30% and 70% shade) experienced a spike in temperatures that exceeded 50°C and 60°C in the winter and summer, respectively (S2 Fig).

Fig 1. A. Shade arena for egg exposure trials showing 99%, 70% and 30% shade sections, and B. egg strips with data logger.

Egg strips were obtained by exposing sandpaper strips to gravid mosquitoes. Eggs were allowed to incubate within the flooded oviposition cup for 3 days, then dried. Viable (non flattened) eggs on strips were counted and each strip labeled with egg count using masking tape on a bulldog clip). Eggs strips were attached vertically to the inside of a wire mesh basket that was set on a card table within each section (Fig 1). The legs of the table were set inside a saucer containing talcum powder to prevent ants from invading the table and predating eggs. Three egg strips were randomly picked and hatched in a dilute yeast solution at 0, 1, 2, 3, 4, 6, 8, 12 and 16 weeks post exposure. The experiment was conducted in the cool dry (10 May– 30 Aug. 2012) and warm dry season (11 Oct. 2012–3 Jan. 2013). Percent hatch was calculated as the number of hatched eggs (1st instar larvae) divided by total viable eggs.

To assess changes in hatch rates across time in the summer and winter egg quiescence experiment, we considered rates observed in the first 6 weeks or less and then ran General Linear Models (GLMs) examining the impact of strain and shade on hatch rates, with time treated as a linear and quadratic variable. We were particularly interested in the interaction between time and strain reflecting changes in hatch rate over time. GLMs were run with and without angularly transforming hatch rates but only the latter are represented because the conclusions were identical.

Elimination trial

We conducted a trial to see if establishing a fixed population of wMelPop infected Ae. aegypti could be used to crash the viable egg bank during the dry season and potentially eliminate the local population under semi-field cage conditions. Two semi-field cages (4 x 7 x 4 m) were populated with wild (F1-2) Cairns Ae. aegypti. Each cage contained 4 10-L buckets and 4 potplants with flooded potplant base (PPB). Three times per week a volunteer sat in the cage for 10 min to bloodfed mosquitoes; 1 pellet of lucerne was added to each bucket and the potplant base was refilled with water. Buckets and PPBs were flooded weekly to hatch new eggs. The populations were introduced into the cage on 28 April 2013 and allowed to establish for 4 weeks. Then populations were crashed (simulated source reduction campaign) by emptying 3 of the 4 buckets and PPBs to kill immatures then spraying the inside of each with 4% sodium hypochlorite (commercial bleach) to kill eggs [30]. One week after population suppression, PPB and buckets were refilled with water. Mean pupal counts were used to estimate the adult population in both cages by integrating estimated adult populations for 0.8 and 0.9 daily survival over 14 days [31,32]. Single Biogents Sentinel traps (BGS) were run for 3 hr in each cage 2–3 times/week to monitor adult populations. We used the relative change in BGS captures during the 2 weeks before and after releases of wMelPop-infected mosquitoes to estimate the size of the released cohort [33].

On 6 June 2013 weekly release of wMelPop infected mosquitoes began in the treatment cage. A colony of AOMB wMelPop cultured at a constant temperature was the source of release material. Two cups of 50 (1:1 male:female) adult Ae. aegypti infected with wMelPop were released into the north (treatment cage). This equated to a relative release ratio of 1.4 and 2.0 for female and male infected mosquitoes, respectively, based on the relative change in BGS captures from 2 weeks before and after release [33]. Mosquito populations were also monitored in both cages by counting pupae in buckets and PPBs once week. Mosquitoes in the control and treatment cage were monitored for Wolbachia infection by PCR analysis of sample of up to 30 larvae/week; larvae were reared to adults for PCR. We also conducted another source reduction (3/4 of buckets and PPBs treated as before) supplemented with removal of adults using a sweepnet on 5 July. Releases of Wolbachia infected mosquitoes continued until 100% of sampled mosquitoes in the treatment cage were infected with wMelPop for two successive weeks; the last release was on 26 Sept. 2013.

We characterized the infection frequency in the cage by testing samples of 30–90 adults (collected as larvae) for the Wolbachia infection. DNA was extracted from adult mosquitoes by using a Chelex 100 resin (Bio-Rad Laboratories, Hercules, CA) extraction method. 38 mosquitoes were ground in 3 L of proteinase K (20 mg/mL) (Roche Diagnostics Australia Pty. Ltd., Castle Hill New South Wales, Australia) and 250 L of 5% Chelex solution, incubated at 65°C for 1 hour, followed by incubation for 10 minutes at 90°C and storage at -20°C. Wolbachia infection status was determined by using methods developed for a Roche LightCycler 480 [34] with a modification to the primers used as described elsewhere [25]. Quantitative real-time polymerase chain reaction (PCR) was used to amplify three markers with three primer sets: a mosquito primer set to detect mosquitoes from the Aedes genus, Ae. aegypti specific primers to differentiate Ae. aegypti from other Aedes species, and Wolbachia specific primers to determine Wolbachia infection status, as well as density. PCR conditions were as follows: 95°C for 10 minutes, 40 cycles of 95°C for 5 seconds, 58°C for 15 seconds and 72°C for 15 seconds, ending with a 95°C 1-minute heating followed by cool down to 40°C for 20 seconds before raising to 65°C. A melting curve analysis was performed via a gradual increase of temperature from 65°C to 95°C. Diagnosis was based on crossing-point values for the PCR and melting temperature from high resolution melt analysis.

We tested the potential for wMelPop to eliminate the field cage population of Ae. aegypti. Mosquitoes in both the treatment and control cages were allowed oviposit for three weeks after the last release. To provide a substrate that could be easily divided into separate egg cohorts to measure survival, we placed a roughened plastic liner inside each bucket for oviposition [30]. Sections of each liner were then cut off with scissors, viable eggs counted microscopically, and placed within the 70% and 99% shade sections of the awning used to measure egg survival. Strips of wMelPop infected and uninfected eggs were hatched in dilute yeast solution after 29, 59 and 80 days exposure (7 Nov, 16 Dec and 6 Jan). Strips were dried for 1 week then reflooded to hatch any viable eggs not initially hatched. Percent hatch was calculated as with the egg survival study. In the final hatch all potplant bases and buckets were flooded to see if any eggs that may have been laid on the container rather than the ovistrip survived. We compared the total percent hatch between treatment and control for exposure period using chi square.

Selection and backcrossing

Selection experiments were conducted to examine if egg desiccation resistance was due to a nuclear or Wolbachia effect. Eggs were collected daily for four days using sandpaper ovistrips and were fully conditioned as per the standard protocol [22]. All ovistrips were placed into a desiccation chamber set at ~60% Relative Humidity (RH) using a saturated solution of sodium bromide [22]. RH was monitored using a hygrochron (1-wire,

To select a wMelPop desiccation resistant line (AOMB-DS), we were aiming to obtain around 10% hatch rate over four rounds of selection. Previous experiments have shown that a 10% hatch rate occurred around Day 20–21. Therefore, the ovistrips were divided into batches and hatched over a range of days (18–26) to obtain a hatch rate close to 10%. Eggs were hatched using Reverse Osmosis (RO) water, Tetramin fish food and yeast [22]. Hatch rate was determined by (number of eggs/number of larvae) x 100. In each generation of selection, at least 100 eggs were used in the next generation. Each round of selection was followed by a non-selected generation to increase the number of adults for producing eggs available for selection in the next generation. There were four rounds of selection and the colony was then maintained in the laboratory over several generations.

To determine if the selection response in AOMB-DS was due to a nuclear or Wolbachia effect, the selected line was crossed to unselected lines. Two new lines (AX1, AX2) were created by backcrossing AOMB-DS to males or females of the original AOMB line. Backcrossing was continued for three generations to place Wolbachia from the selected line onto the nuclear background of the infected (AX1) stock (Fig 2) or to place Wolbachia from the unselected line onto the nuclear background of the selected infected line (AX2).

Fig 2. Crossing scheme to produce colonies with the selected Wolbachia background (AX1) or selected nuclear background (AX2).

To test for viability, eggs were collected over four day periods from the AOMB-DS, AOMB, and C20 colonies as well as the backcrossed AX1 and AX2 lines. These were placed in the 60% desiccation chamber once conditioned. Three days later, eggs were cut into 10 batches of at least 25 eggs then counted and hatched into separate cups filled with 170 ml RO water and TetraMin fish food. Larvae were counted one week later as an estimate of viability. This process was carried out on eggs aged for 3 days and then repeated after 10, 17, 24, 31, 38, 45, 52, 59 and 66 days [22].

Following selection, to compare the association between hatch rates and time in the different strains, we truncated the changes in hatch rates at 32 days (ie before all the unselected wMelPop eggs had become non-viable) and ran GLMs that included linear and quadratic terms for time as well as line type. These parameters provided an adequate fit to the data or each line (R2>0.90). We then considered whether lines differed for estimated parameters in these models. Because multiple strains were compared we computed parameter estimates for the different treatments and interactions and tested their significance using bootstrapped confidence intervals estimated in IBM SPSS Statistics 22.

To determine if the selection response in AOMB-DS was due to a nuclear or Wolbachia effect, the selected line was crossed to unselected lines. Two new lines (AX1, AX2) were created by backcrossing AOMB-DS to males or females of the original AOMB line. Backcrossing was continued for three generations to place Wolbachia from the selected line onto the nuclear background of the infected (AX1) stock (Fig 2) or to place Wolbachia from the unselected line onto the nuclear background of the selected infected line (AX2).

To test for viability, eggs were collected over four day periods from the AOMB-DS, AOMB, and C20 colonies as well as the backcrossed AX1 and AX2 lines. These were placed in the 60% desiccation chamber once conditioned. Three days later, eggs were cut into 10 batches of at least 25 eggs then counted and hatched into separate cups filled with 170ml RO water and TetraMin fish food. Larvae were counted one week later as an estimate of viability. This process was carried out on eggs aged for 3 days and then repeated after 10, 17, 24, 31, 38, 45, 52, 59 and 66 days [22].

Data repository

All summary data can be accessed in Dryad Digital Repository at [35].


Egg survival

The wMelPop infected Ae. aegypti eggs exposed to a range of shade (Fig 1) suffered higher mortality rates than uninfected eggs (Fig 3). Mortality (decrease in % hatch) was highest in summer and in sun exposed (30%, 70% shade) eggs. Indeed, mortality of wMelPop infected eggs was nearly 100% after 8 weeks in all shade groups in summer. Uninfected eggs also suffered high mortality, but only in the highly exposed 30% shade treatment. The GLM on data from the summer experiment indicated a significant effect of shade and strain overall as well as time as a linear and quadratic term (Table 1). Interactions between strain and the time variables were not significant, although the quadratic—strain interaction approached significant. Increased shade led to lower viability across time (Fig 3) and the parameter estimates indicated significant differences between the 99% shade and the 30% treatment, but there was no interaction between shade and strain or other interactions with shade (all P>0.25). The wMelPop strain had lower viability overall and tended to show a more rapid decrease in viability over time.

Table 1. Effect of shade and strain on hatch rates (angular transformed) of eggs across weeks (evaluated as linear and quadratic components).

Fig 3. Hatch rate (survival) of Ae. aegypti eggs infected with wMelPop and an uninfected control held in 30%, 70% and 99% shade in winter and summer.

For the winter experiment, the strain-shade interaction effects were also non-significant (P > 0.25). All other effects except for the main effect of strain were highly significant (Table 1). There were different patterns of changes in the hatch rates of strains across weeks, with linear and quadratic components. The wMelPop infection had a sharper drop in hatch rates compared to the uninfected line (Fig 3). The temperature in the winter trial ranged from ca. 15–28°C, with slightly higher temperatures in the 30% shade section (S1 Fig). In winter, temperatures recorded from loggers on tables exposed to direct sunlight (30% and 70% shade) peaked at 40–50°C in the afternoon (S2 Fig). Relative humidity was generally high, ranging from ca. 50% to nearly 100% at night, although the maximum values are likely too high (S3 Fig). In summer, temperature range increased to ca. 20–35°C (S4 Fig), with exposed loggers peaking at 40–60°C at mid day for 30% and 70% shade (S2 Fig). Relative humidity ranged from 50% to 90% (S3 Fig).

Elimination trial

Releases resulted in the wMelPop infection gradually moving to fixation within the semi-field cage. Two successive weeks of 100% wMelPop infection rates were achieved after 16 consecutive weekly releases and two source reduction/vector control interventions (Fig 4).

Fig 4. Frequency of infection in the semi-field cage with error bars representing 95% confidence intervals.

Weekly releases of 50 male/female wMelPop infected mosquitoes began on 6 June and continued through 23 Sept. Vector control consisting of treatment of 75% of breeding containers was conducted on 30 May and 5 July. Adult mosquitoes (> 40 sample) reared from larvae/pupae were assayed for wMelPop infection by PCR.

The mean temperature in the 70% and 99% shade sections was 26.2°C and 25.4°C, respectively, and ranged from 20–35°C (S5 Fig). Relative humidity ranged from 50% to 95%. While we did not record temperatures in the exposed data loggers (set on tables), results from the earlier egg survival study indicate that temperatures of 50°C would have occurred in the 70% shade section for a few hours on sunny days (S2 Fig). Mean (± SE) pupal counts for individual PPB and buckets in the wMelPop and uninfected cages were 0.62 ± 0.96 and 1.06 ± 1.37; and 28.9 ± 30.6 and 50.4 ± 27.2, respectively. Collectively, the mean pupal production/day was 59 ± 49 and 103 ± 46 for the wMelPop and control cages, respectively. Based on a 0.8 and 0.9 adult daily survival (DS), this would equate to a mean adult population, using the integration method [30], of 280–451 and 490–790 adults for the wMelPop and uninfected cages, respectively.

Eggs infected with wMelPop suffered significantly higher mortality than uninfected eggs for each exposure period except for 80 days at 70% shade, where no eggs hatched for either cohort (Fig 5). wMelPop infected egg survival dropped noticeably after 59 days, and only 3/443 (0.67%) eggs hatched from the 99% shade cohort after 80 days. Mortality was also high in the uninfected eggs, but 7/154 (4.4%) eggs hatched in the 99% shade cohort. No wMelPop infected eggs hatched from the PPB and buckets flooded at the end of the trial, while 9 hatched from a PPB with uninfected eggs.

Fig 5. Egg survival (egg hatch) of wild and wMelPop infected Ae. aegypti eggs held outdoors in A. 70% and B. 99% shade from 18 Oct. 2013–6 Jan. 2014.

** and **** denote P < 0.01 and P < 0.0001, respectively.

Selection and backcrossing

After four rounds of selection for desiccation resistant mosquitoes, there was an increase in the time required to produce a hatch rate of around 10% from around 17 days to almost 40 days (Table 2). This followed a rapid extension after two rounds of selection.

Table 2. Response to selection for desiccation resistance through four rounds of selection.

In the backcrossed lines, the AOMB-DS line maintained a higher hatch rate in aged eggs compared to the AOMB line despite several generations without selection (Fig 6). The AX2 line showed a similar pattern to that of the AOMB-DS line. As expected, the uninfected C20 had the highest hatch rate. The unselected AOMB and AX1 lines had similar low hatch rates. The GLM showed that there was a significant effect of line on hatch rate (F(4, 285) = 4.097, P = 0.003) and there were also significant interactions between line and the linear time component (F(4,285) = 6.974, P<0.001) and between line and the quadratic time component (F(4,285) = 5.042, P = 0.001). These analyses indicate differences among strains in hatch rate changes across time involving both a linear and non-linear component. The time effects point to an increase in the hatch rate (survival) of the selected line with the wMelPop infection associated with the nuclear background of the line rather than the Wolbachia background. The AX1 line and AOMB did not differ in any way and all parameter estimates overlapped for these strains (main effects of strain, interactions with linear and quadratic time components). This was also the case for AX2 and AOMB-DS. In contrast, the linear parameter for AX2 and week (-0.103) fell outside the 95% confidence intervals for the equivalent value for AOMB (-0.582, -0.192) and this was also the case for the quadratic component (0.001) which fell outside the 95% confidence intervals for AOMB (0.015, 0.101). The quadratic component also differed between the selected and unselected AOMB lines (AOMB, 0.058, 95% CIs 0.015, 0.101; AOMB-DS, -0.031, 95% -0.074, 0.013) as did the linear component (AOMB, -0.387, 95% CIs -0.582, -0.192; AOMB-DS, 0.051, 95% -0.144, 0.246).

Fig 6. Changes in the hatch rates of the lines when eggs are maintained in a quiescent stage for different lengths of time before being hatched.

An uninfected line (C20; top line) was included as a control. The other lines are the infected wMelPop line (AOMB), the same line after four rounds of selection (AOMB-DS), and backcross lines to place the wMelPop infection from the selected line onto the nuclear background of the unselected line (AX1) or the infection from the unselected line onto the nuclear background of the selected line (AX2).


Aedes aegypti eggs infected with wMelPop suffer significantly higher mortality than uninfected eggs. Nearly all of the infected eggs failed to hatch after 2 months for all shade and temperature exposures (Fig 2). Our semi-field cage experiment indicted that if releases of wMelPop successfully establish fixation, the population will potentially collapse during prolonged dry periods. Indeed, the hatching rate (survival) of wMelPop infected eggs was significantly lower after 59 and 80 days exposure (Fig 3). This points to the feasibility of using releases of wMelPop infected Ae. aegypti to eliminate local populations of Ae. aegypti and confirms the potential of this approach as first proposed and modelled in Rasic et al. [26].

Nevertheless, using wMelPop releases for elimination of Ae. aegypti populations will be more difficult than invasions involving wMel releases. To obtain fixation in the semi-field cage, we conducted 16 weeks of releases and two rounds of vector control. This is considerably more difficult than the invasions of wMel that have successfully established in Cairns after 11 weeks of releases, and at a faster rate in semi-field cages [17,19]. Indeed, sustained releases of wMelPop-infected Ae. aegypti failed to establish fixation by wMelPop at two field sites in Cairns in 2012 [25]. We could have improved efficacy of the releases by increasing the number of infected mosquitoes released. At the start of releases, the population in the treatment cage, based on pupal integration, was ca. 200 adults. We released 100 per week, or a release rate equivalent to 50% of uninfected adult population, although BGS captures suggested the initial 2 weeks of release cohorts were 150% of the uninfected population. The successful wMel invasions in Cairns involved releases at numbers roughly equivalent to wild populations. Higher releases rates should have improved the speed at which wMelPop became fixed in our semi-field cage.

The use of wMelPop to eliminate local populations of Ae. aegypti will likely be limited by several constraints. It took 16 weeks of releases, with 2 cycles of vector control, to fix wMelPop in a small semi-field cage. Thus, releases without vector control, population suppression or another mechanism to assist invasion like the use of insecticide resistance are likely to be challenging [36]. The site will likely need to be isolated to avoid rapid re-infestation from adjacent areas, and will likely need to be relatively small, such as an isolated township [26]. Also the locale must have a prolonged dry period to facilitate egg death, as modelled in Rasic et al. [26]. Indeed, there were 4 viable infected eggs (3 from a bucket, and 1 from a PPB) after 80 days in the full shade cohort, suggesting that wMelPop may persist at a very low level requiring additional vector control for elimination. Finally, candidate areas need to be focused on the elimination of Ae. aegypti, rather than simply a reduction of dengue risk, which will be more easily achieved by releases of wMel rather than wMelPop. The data generated from this study could be used to identify geographic areas through modelling where the use of virulent Wolbachia and vector control for population suppression would be most effective.

Vector control and population manipulation can be used to facilitate wMelPop releases [36]. Improved vector control methods that target cryptic breeding sites, such as the use of auto-dissemination application of pyriproxyfen [37], could temporarily remove cryptic foci of uninfected mosquitoes, enhancing invasion. Other temporary vector control methods, such as use of bleach to treat larval habitat [30] and non-persistent adulticides such as metofluthrin [38], could provide a quick knockdown and avoid residual effects that would impact released mosquitoes (crash and release strategy). A novel approach would be to couple genes for pesticide resistance to wMelPop, in effect engendering protection of the wMelPop mosquitoes from insecticides [39], although this approach may be coupled with the fear of introducing resistant strains, and could be difficult to sell to the public, regulators and government officials.

While the wMelPop infection has deleterious effects on egg viability when these are in a dried state, these effects are attenuated by strong selection. The crosses point to this attenuation being due to a nuclear background effect (Fig 6). Previous research has shown that the expression of Wolbachia effects on the host can be influenced by host background. Perhaps the most dramatic example of this is where Wolbachia causing cytoplasmic incompatibility or male killing in one host and then transfected into a different host species no longer expresses the same phenotype [40]. In addition, lifespan effects generated by Wolbachia in Drosophila melanogaster are altered through longevity selection and involve the nuclear background [41]. Once wMelPop have been successfully introduced into populations, evolution in Wolbachia may attenuate fitness effects, as recently documented in Drosophila [42] although this is clearly not always the case. Evolution seems likely to change the nuclear background and lead to attenuation, but this process may be countered by other selective pressures.

Does use of a virulent Wolbachia strain such as wMelPop to eliminate local Ae. aegypti populations have advantages over traditional releases of sterile (sterile insect technique (SIT)) or incompatible (incompatible insect technique (IIT)) males? Certainly use of Wolbachia eliminates the need for costly irradiation equipment and any potential fitness cost associated with irradiating males. Reared Wolbachia-infected cohorts do not need to be sexed before release, and the relative numbers of released insects should be much lower that SIT programs where overwhelming release ratios of high quality males are required [43]. There is no sudden population “bounce back” if releases of Wolbachia infected mosquitoes stopped (assuming Wolbachia is nearly fixed in the population) unlike SIT/IIT where residual wild populations can rapidly rebound. Wolbachia is seen as a biological control program with generally strong community support [44], whereas some IIT programs involve genetically modified organisms, and may be relatively difficult to obtain public and regulatory support for. Disadvantages of use of virulent Wolbachia include the release of biting females, of the need to create a suitable infected line, and the potential requirement to conduct complimentary vector control.

The survival of uninfected Ae. aegypti eggs exposed to direct sun (30 and 70% shade) deserves comment. While egg survival was highest in full shade, many eggs exposed to sunlight (30% and 70% shade) survived several weeks (Fig 2). Ae. aegypti eggs exposed to temperatures > 46°C suffer rapid mortality [45,46]. Our data loggers recorded mid day temperature spikes of up to 40°C to 60°C, 20–30°C higher than ambient temperature, lasting several hours (S2 Fig). We suspect that the loggers, oriented horizontally on top of the sample jar (Fig 1), were subject to high solar gain and heating, and thus misleadingly high temperature readings. The Aedes eggs, on the other hand, are oriented vertically, minimizing their silhouette area when the sun is overhead and thus reducing solar gain. The small size of eggs would also facilitate cooling by the air.

In summary, we have established the effects of the wMelPop infection on egg mortality when these are maintained in a dried state under semi-field conditions. This provides the possibility of using wMelPop to suppress and even eliminate mosquito populations in relatively isolated populations particularly where there is a long dry season.

Supporting Information

S1 Fig. Temperature (daily high, low and average from logger in shade) in 30%, 70% and 99% shade sections of the egg exposure unit in winter.


S2 Fig. Temperature from data loggers exposed to sunlight in egg exposure unit in winter and summer.


S3 Fig. Relative humidity in 99% shade section of the egg exposure unit during the egg survival experiment.


S4 Fig. Temperature (daily high, low and average from logger in shade) in 30%, 70% and 99% shade sections of the egg exposure unit in summer.


S5 Fig. Temperature and relative humidity (99% shade) in 70% and 99% section of egg exposure unit during the wMelPop elimination trial.



We thank Warren Porter of University of Wisconsin for reviewing the manuscript and for advice on biophysical aspects of solar exposure of mosquito eggs and data loggers, and Herve Bossin of Institute Louis Malarde for comments on the manuscript. We thank Brian Johnson of JCU for helping prepare figures.

Author Contributions

Conceived and designed the experiments: SAR AAH CP. Performed the experiments: SAR MT CP. Analyzed the data: SAR AC AAH. Contributed reagents/materials/analysis tools: AC AAH. Wrote the paper: SAR AAH.


  1. 1. Bhatt S, Gething PW, Brady OJ, Messina JP, Farlow AW, Moyes CL et al. The global distribution and burden of dengue. Nature. 2013; 496(7446): 504–507. pmid:23563266
  2. 2. Flores AE, Ponce G, Silva BG, Gutierrez SM, Bobadilla C, Lopez B et al. Widespread cross resistance to pyrethroids in Aedes aegypti (Diptera: Culicidae) from Veracruz state Mexico. J Econ Entomol. 2013; 106: 959–969. pmid:23786088
  3. 3. Koou SY, Chong CS, Vythilingam I, Ng LC, Lee CY. Pyrethroid resistance in Aedes aegypti larvae (Diptera: Culicidae) from Singapore. J Med Entomol. 2014; 51: 170–181. pmid:24605467
  4. 4. Linss JG, Brito LP, Garcia GA, Araki AS, Bruno RV, Lima JB, et al. Distribution and dissemination of the Val1016Ile and Phe1534Cys Kdr mutations in Aedes aegypti Brazilian natural populations. Parasit Vectors. 2014; 7: 25. pmid:24428880
  5. 5. Maciel-de-Freitas R, Avendanho FC, Santos R, Sylvestre G, Araújo SC, Lima JB, et al. Undesirable consequences of insecticide resistance following Aedes aegypti control activities due to a dengue outbreak. PLoS One. 2014; 9: e92424. pmid:24676277
  6. 6. Barrera R, Amador M, Diaz A, Smith J, Munoz-Jordan JL, Rosario Y. Unusual productivity of Aedes aegypti in septic tanks and its implications for dengue control. Med Vet Entomol. 2008; 22: 62–69. pmid:18380655
  7. 7. Russell BM, Mcbride WJ, Mullner H, Kay BH. Epidemiological significance of subterranean Aedes aegypti (Diptera: Culicidae) breeding sites to dengue virus infection in Charters Towers, 1993. J Med Entomol. 2002; 39: 143–145. pmid:11931248
  8. 8. Montgomery BL, Ritchie SA. Roof gutters: a key container for Aedes aegypti and Ochlerotatus notoscriptus (Diptera: Culicidae) in Australia. Am J Trop Med Hyg. 2002; 67: 244–246. pmid:12408662
  9. 9. Reiter P, Gubler DJ. Surveillance and control of urban dengue vectors. In: Kuno G, editors. Dengue and dengue hemorrhagic fever. CAB International New York. 1997; pp. 45–60.
  10. 10. Perich MJ, Davila G, Turner A, Garcia A, Nelson M. Behavior of resting Aedes aegypti (Culicidae: Diptera) and its relation to ultra-low volume adulticide efficacy in Panama City, Panama. J Med Entomol. 2000; 37: 541–546. pmid:10916294
  11. 11. Alphey L, Benedict M, Bellini R, Clark GG, Dame DA, Service MW, et al. Sterile-insect methods for control of mosquito-borne diseases: an analysis. Vector Borne Zoonotic Dis. 2010; 10: 295–311. pmid:19725763
  12. 12. Harris AF, Nimmo D, McKemey AR, Kelly N, Scaife S, Donnelly CA, et al. Field performance of engineered male mosquitoes. Nat Biotechnol. 2011; 29: 1034–1037. pmid:22037376
  13. 13. McMeniman CJ, Lane RV, Cass BN, Fong AW, Sidhu M, Wang YF, et al. Stable introduction of a life-shortening Wolbachia infection into the mosquito Aedes aegypti. Science. 2009; 323: 141–144. pmid:19119237
  14. 14. McGraw EA, O'Neill SL. Beyond insecticides: new thinking on an ancient problem. Nat Rev Microbiol. 2013; 11: 181–193. pmid:23411863
  15. 15. Turelli M. Cytoplasmic incompatibility in populations with overlapping generations. Evolution. 2010; 64: 232–241. pmid:19686264
  16. 16. Barton NH, Turelli M. Spatial waves of advance with bistable dynamics: cytoplasmic and genetic analogues of Allee effects. Am Nat. 2011; 178: E48–E75. pmid:21828986
  17. 17. Hoffmann AA, Montgomery BL, Popovici J, Iturbe-Ormaetxe I, Johnson PH, Muzzi F, et al. Successful establishment of Wolbachia in Aedes populations to suppress dengue transmission. Nature. 2011; 476: 454–457. pmid:21866160
  18. 18. Moreira LA, Iturbe-Ormaetxe I, Jeffery JA, Lu G, Pyke AT, Hedges LM, et al. A Wolbachia symbiont in Aedes aegypti limits infection with dengue, Chikungunya, and Plasmodium. Cell. 2009; 139: 1268–1278. pmid:20064373
  19. 19. Walker T, Johnson PH, Moreira LA, Iturbe-Ormaetxe I, Frentiu FD, McMeniman CJ, et al. The wMel Wolbachia strain blocks dengue and invades caged Aedes aegypti populations. Nature. 2011; 476: 450–453. pmid:21866159
  20. 20. van den Hurk AF, Hall-Mendelin S, Pyke AT, Frentiu FD, McElroy K, Day A, et al. Impact of Wolbachia on infection with chikungunya and yellow fever viruses in the mosquito vector Aedes aegypti. PLoS Negl Trop Dis. 2012; 6: e1892. pmid:23133693
  21. 21. McMeniman CJ, O'Neill SL. A virulent Wolbachia infection decreases the viability of the dengue vector Aedes aegypti during periods of embryonic quiescence. PLoS Negl Trop Dis. 2010; 4: e748. pmid:20644622
  22. 22. Yeap HL, Mee P, Walker T, Weeks AR, O'Neill SL, Johnson P, et al. Dynamics of the "popcorn" Wolbachia infection in outbred Aedes aegypti informs prospects for mosquito vector control. Genetics. 2011; 187: 583–595. pmid:21135075
  23. 23. Turley AP, Moreira LA, O'Neill SL, McGraw EA. Wolbachia infection reduces blood-feeding success in the dengue fever mosquito, Aedes aegypti. PLoS Negl Trop Dis. 2009; 3: e516. pmid:19753103
  24. 24. Ross PA, Endersby NM, Yeap HL, Hoffmann AA. Larval competition extends developmental time and decreases adult size of wMelPop Wolbachia-infected Aedes aegypti. Am J Trop Med Hyg. 2014; 91:198–205. pmid:24732463
  25. 25. Yeap HL, Axford JK, Popovici J, Endersby NM, Iturbe-Ormaetxe I, Ritchie SA, et al. Assessing quality of life-shortening Wolbachia-infected Aedes aegypti mosquitoes in the field based on capture rates and morphometric assessments. Parasit Vectors. 2014; 7: 58. pmid:24495395
  26. 26. Rašić G, Filipović I, Weeks AR, Hoffmann AA. Genome-wide SNPs lead to strong signals of geographic structure and relatedness patterns in the major arbovirus vector, Aedes aegypti. BMC Genomics. 2014; 15: 275. pmid:24726019
  27. 27. Ritchie SA (2014) Dengue vector bionomics: Why Aedes aegypti is such a good vector. Ch. 24 in Dengue and Dengue Hemorrhagic Fever. Gubler D, Ooi EE, Vasudevan S, Farrar J. (Eds) CAB International, Oxfordshire UK. 2014.
  28. 28. Segoli M, Hoffmann AA, Lloyd J, Omodei GJ, Ritchie SA. The effect of virus-blocking Wolbachia on male competitiveness of the dengue vector mosquito, Aedes aegypti. PLoS Negl Trop Dis. 2014; 8: e3294. pmid:25502564
  29. 29. Ritchie SA, Johnson PH, Freeman AJ, Odell RG, Graham N, Dejong PA, et al. A secure semi-field system for the study of Aedes aegypti. PLoS Negl Trop Dis. 2011; 5: e988. pmid:21445333
  30. 30. Jacups SP, Ball TS, Paton CJ, Johnson PH, Ritchie SA. Operational use of household bleach to "crash and release" Aedes aegypti prior to Wolbachia-infected mosquito release. J Med Entomol. 2013; 50: 344–351. pmid:23540123
  31. 31. Focks DA, Sackett SR, Bailey DL, Dame DA. Observations on container-breeding mosquitoes in New Orleans, Louisiana, with an estimate of the population density of Aedes aegypti (L.). Am J Trop Med Hyg. 1981; 30: 1329–1335. pmid:7325289
  32. 32. Williams CR, Johnson PH, Ball TS, Ritchie SA. Productivity and population density estimates of the dengue vector mosquito Aedes aegypti (Stegomyia aegypti) in Australia. Med Vet Entomol. 2013; 27: 313–322. pmid:23205694
  33. 33. Ritchie SA, Montgomery BL, Hoffmann AA. Novel estimates of Aedes aegypti population size and adult survival based on Wolbachia releases. J Med Entomol. 2013; 50: 624–631. pmid:23802459
  34. 34. Lee SF, White VL, Weeks AR, Hoffmann AA, Endersby NM. High-throughput PCR assays to monitor Wolbachia infection in the dengue mosquito (Aedes aegypti) and Drosophila simulans. Appl Environ Microbiol. 2012; 78: 4740–4743. pmid:22522691
  35. 35. Ritchie SA, Townsend M, Paton CJ, Callahan AG, Hoffmann AA (2015) Data from: Application of wMelPop Wolbachia strain to crash local populations of Aedes aegypti. Dryad Digital Repository.
  36. 36. Hoffmann AA. Facilitating Wolbachia invasions. Austral J Entomol. 2014; 53: 125–132.
  37. 37. Devine GJ, Perea EZ, Killeen GF, Stancil JD, Clark SJ, Morrison AC. Using adult mosquitoes to transfer insecticides to Aedes aegypti larval habitats. Proc Natl Acad Sci U S A. 2009; 106: 11530–11534. pmid:19561295
  38. 38. Ritchie SA, Devine GJ. Confusion, knock-down and kill of Aedes aegypti using metofluthrin in domestic settings: a powerful tool to prevent dengue transmission? Parasit Vectors. 2013; 6: 262. pmid:24025232
  39. 39. Hoffmann AA, Turelli M. Facilitating Wolbachia introductions into mosquito populations through insecticide-resistance selection. Proc Biol Sci. 2013; 280: 20130371. pmid:23576788
  40. 40. Veneti Z, Zabalou S, Papafotiou G, Paraskevopoulos C, Pattas S, Livadaras I, et al. Loss of reproductive parasitism following transfer of male-killing Wolbachia to Drosophila melanogaster and Drosophila simulans. Heredity. 2012; 109: 306–312. pmid:22892635
  41. 41. Carrington LB, Hoffmann AA, Weeks AR. Monitoring long-term evolutionary changes following Wolbachia introduction into a novel host: the Wolbachia popcorn infection in Drosophila simulans. Proc Biol Sci. 2010; 277: 2059–2068. pmid:20219740
  42. 42. Chrostek E, Teixeira L. Mutualism breakdown by amplification of Wolbachia genes. PLoS Biology. 2015; 13: e1002065. pmid:25668031
  43. 43. Van der Vloedt AM, Klassen W. The development and application of the sterile insect technique (SIT) for New World screwworm eradication. Rome, Italy: Food and Agriculture Organization World Animal Review. 1991.
  44. 44. Kolopack PA, Parsons JA, Lavery JV. What makes community engagement effective?: lessons from the Eliminate Dengue Program in Queensland Australia. PLoS Negl Trop Dis. 2015; 9(4): e0003713. pmid:25875485
  45. 45. Christophers SR. Aedes aegypti (L.) The yellow fever mosquito. London: Cambridge University Press. 1960.
  46. 46. Smith GC, Eliason DA, Moore CG, Ihenacho EN. Use of elevated temperatures to kill Aedes albopictus and Ae. aegypti. J Am Mosq Control Assoc. 1988; 4: 557–558. pmid:3225579