Continued Susceptibility of the wMel Wolbachia Infection in Aedes aegypti to Heat Stress Following Field Deployment and Selection

Aedes aegypti mosquitoes infected with the wMel strain of Wolbachia are being deployed to control the spread of arboviruses around the world through blockage of viral transmission. Blockage by Wolbachia in some scenarios may be affected by the susceptibility of wMel to cyclical heat stress during mosquito larval development. We therefore evaluated the potential to generate a heat-resistant strain of wMel in Ae. aegypti through artificial laboratory selection and through exposure to field temperatures across multiple generations. To generate an artificially selected strain, wMel-infected females reared under cyclical heat stress were crossed to wMel-infected males reared at 26 °C. The low proportion of larvae that hatched founded the next generation, and this process was repeated for eight generations. The wMel heat-selected strain (wMel-HS) was similar to wMel (unselected) in its ability to induce cytoplasmic incompatibility and restore compatibility when larvae were reared under cyclical heat stress, but wMel-HS adults exhibited reduced Wolbachia densities at 26 °C. To investigate the effects of field exposure, we compared the response of wMel-infected Ae. aegypti collected from Cairns, Australia where the infection has been established for seven years, to a wMel-infected population maintained in the laboratory for approximately 60 generations. Field and laboratory strains of wMel did not differ in their response to cyclical heat stress or in their phenotypic effects at 26 °C. The capacity for the wMel infection in Ae. aegypti to adapt to high temperatures therefore appears limited, and alternative strains may need to be considered for deployment in environments where high temperatures are regularly experienced in mosquito breeding sites.


Introduction
The open field release of Wolbachia-infected mosquitoes is a promising approach to control the spread of arboviruses and has gained considerable attention in recent years [1][2][3]. Wolbachia infections can protect their hosts against pathogens, and this feature has been exploited to generate strains of mosquitoes with a reduced ability to transmit arboviruses [4,5]. The principal vector of dengue, the Aedes aegypti mosquito, has now been infected experimentally with several strains of Wolbachia originating from other insects [4,[6][7][8][9][10][11]. The majority of these Wolbachia strains induce cytoplasmic incompatibility, which greatly reduces or eliminates the production of viable offspring by uninfected females that mate with Wolbachia-infected males. This feature combined with Wolbachia-induced viral suppression has prompted the deployment of Ae. aegypti with artificial Wolbachia infections into the field in areas where dengue and other arboviruses are endemic. Efforts such as the World Mosquito Program (formerly Eliminate Dengue) (www.worldmosquitoprogram.org) and Wolbachia Malaysia (http://repository.imr.gov.my/wolbachia) aim to replace natural populations of Ae. aegypti mosquitoes proteins [41]. These studies demonstrate that Wolbachia infections can evolve in response to a changing environment and raise the possibility that experimental evolution or direct selection could therefore be used as a tool to select Wolbachia infections in Ae. aegypti mosquitoes for desirable attributes, improving their utility for arbovirus control applications.
In this study, we subjected Ae. aegypti infected with wMel to cyclical heat stress over successive generations in an attempt to improve the thermal tolerance of this Wolbachia infection. We also tested whether variable field temperatures had led to any shift in the thermal responses of Wolbachia by comparing a laboratory line exposed to a constant temperature with mosquitoes derived from the field population where the original releases of wMel had taken place seven years ago [1].

Ethics Statement
Blood feeding on human subjects was approved by the University of Melbourne Human Ethics Committee (approval 0723847). All volunteers provided informed written consent.

Mosquito Strains and Colony Maintenance
Ae. aegypti mosquitoes with the wMel infection were collected in 2013 from locations near Cairns, Australia where wMel had successfully established [1]. Uninfected Ae. aegypti were collected in 2016 from locations where wMel-infected mosquitoes had not been released. The wAlbB infection was used for comparisons in some cytoplasmic incompatibility experiments and is described in Xi, Khoo, and Dobson [6] and Axford, et al. [42]. Wolbachia-infected mosquitoes were crossed to uninfected mosquitoes for at least three consecutive generations to ensure similar genetic backgrounds before use in experiments. Colonies were maintained in the laboratory at 26 ± 1 • C according to Ross, et al. [43]. wMel-infected Ae. aegypti were also collected from Yorkeys Knob, Australia in January 2018, where the wMel infection is established in the population following initial releases there seven years ago [1]. At the end of the study, the wMel infection from mosquitoes collected in 2013 had been maintained in the laboratory for approximately 60 generations, while the heat selected wMel population had been through eight generations of selection. These three populations were compared in a later set of experiments.

Selection Regime
We subjected infected Ae. aegypti to cyclical heat stress over successive generations. First instar wMel-infected larvae were placed in incubators (PG50 Plant Growth Chambers, Labec Laboratory Equipment, Marrickville, NSW, Australia) set to cycle diurnally between 26 • C and 37 • C, with 12 h at each temperature, according to Ross, et al. [28]. Up to four trays with 100 larvae each in 500 mL of RO (reverse osmosis) water were reared under these conditions each generation. Data loggers (Thermochron; 1-Wire, iButton.com, Dallas Semiconductors, Sunnyvale, CA, USA) in zip-lock bags were placed in each tray to monitor temperature. Larvae were provided with TetraMin ® tropical fish food tablets (Tetra, Melle, Germany) ad libitum until reaching pupation. Females were separated from males at the pupal stage and returned to 26 • C for adult emergence, and males were discarded. Adult females were held at 26 • C for three days before being crossed to wMel-infected males (unselected) reared at 26 • C. Females were blood-fed on a single human volunteer, and all eggs were hatched in 3 L trays of RO water with a few grains of yeast. The resulting progeny founded the next generation, and the selection regime was repeated. Subsequent gonotrophic cycles were initiated if more eggs were needed.
Crosses with wMel-infected males reared at 26 • C were chosen to produce low hatch proportions, resulting from partial cytoplasmic incompatibility. This increased the intensity of selection as only females that restored compatibility when reared under heat stress would contribute to the next generation, eliminating any individuals that lost their Wolbachia infection. Hatch proportions also Insects 2018, 9, 78 4 of 13 provided an indicator of the response to selection; high hatch proportions in subsequent generations would suggest an improvement in heat tolerance.
After each generation of selection, a subset of progeny was reared at 26 • C as a backup in case the Wolbachia infection from the selected line was lost during the next selection event. After two generations of selection, all progeny were reared at 26 • C to increase the population size as hatch proportions were low (Table 1). After five generations of selection, a subset of progeny was reared at 26 • C for two generations, where females from the selected lines were crossed to wMel (unselected) males. The resulting progeny were then used for cytoplasmic incompatibility experiments and measured for Wolbachia density. Egg hatch proportions were determined by observation under a dissecting microscope and were recorded for the first eight generations. Hatched eggs had a clearly detached cap and no larva inside. We continued selection for three further generations and performed an additional set of experiments that included a wMel infection that had been maintained in the laboratory at 26 • C for approximately 60 generations (wMel-lab), and a wMel infection that had been in the field for seven years (wMel-field). Table 1. Average maximum temperatures and egg hatch rates of wMel-infected Ae. aegypti females for the first eight generations during the selection experiment. Females reared at the specified temperature were crossed to wMel-infected males reared at 26 • C. Generation 3 is a relaxed generation. Only eggs and egg hatch proportions from cages that contributed to the next generation are shown. Egg hatch rates were not recorded after the eighth generation.

Cytoplasmic Incompatibility and Restoration of Compatibility
After five generations of selection followed by two relaxed generations, we tested the ability of the selected wMel strain (from now on referred to as wMel-HS) to induce cytoplasmic incompatibility in crosses with uninfected females and restore compatibility in crosses with wMel-infected males. Larvae from the wMel (unselected), wMel-HS, and wAlbB lines were reared at a constant 26 • C or in incubators set to cycle diurnally between 26 • C and 37 • C, with 12 h at each temperature. Uninfected larvae were reared at 26 • C only for crosses with infected males. We generated two heat stress treatments which differed in the maximum daily temperature reached by 1 • C by utilizing variation within the incubator. Trays were reared on different shelves of the same incubator, reaching average maximum daily temperatures of 36.5 • C and 37.5 • C, respectively ( Figure S1A). Variation between containers on the same shelf was negligible, where they were cycled in their position daily. After five days, larvae from the heat stressed treatments were returned to 26 • C for pupation and adult emergence. Four trays with 100 larvae each were reared for each line and each temperature treatment. Adults from each sex, temperature, and infection status were maintained separately at 26 • C for three days. We then set up nine crosses to test the ability of males from each infected line under each rearing condition to induce cytoplasmic incompatibility with uninfected females. We set up a further nine crosses with females from each infected line under each rearing condition to test their ability to restore compatibility with infected males. For crosses with females from the wMel and wMel-HS lines, males from the unselected wMel line were used. For crosses with wAlbB females, wAlbB males were used. We added 25 males and 25 females to 1.5 L cages for each cross and allowed two days to mate. After blood feeding, 20 females from each cross were isolated in 70 mL cups containing 20 mL of larval rearing water and that were lined with sandpaper for oviposition. Eggs were collected from each cup four days after blood feeding, partially dried, then hatched four days after collection. After one week, females were blood fed again, and eggs were collected and hatched from a second gonotrophic cycle.
After eight generations of selection followed by two relaxed generations, we conducted a similar set of experiments where we included a wMel-infected population maintained in the laboratory for approximately 60 generations (wMel-lab) and a population collected from the field where the wMel infection had been present for seven years (wMel-field). wMel-field was maintained in the laboratory for two generations before experiments began. In this experiment, we reared larvae at 26 • C or under heat stress treatments reaching average maximum daily temperatures of 36 • C and 37 • C ( Figure S1B). Methods for rearing and crosses were otherwise identical to the first experiment, except we tested egg hatch proportions from a single gonotrophic cycle only.

Wolbachia Quantification
A subset of adults from both experiments were stored in absolute ethanol within 24 h of emergence and measured for their Wolbachia density. DNA was extracted from twenty adults from each sex and infection, and Wolbachia density was estimated with qPCR using previously described methods [28,44].

Statistical Analysis
All data were analysed in SPSS statistics version 24.0 for Windows (SPSS Inc., Chicago, IL, USA). Egg hatch proportions and Wolbachia densities were compared with nonparametric Kruskal-Wallis and Mann-Whitney U tests, as data were not normally distributed according to Shapiro-Wilk tests. We used a general linear model to test for an effect of female age on egg hatch proportions during the selection experiment.

Egg Hatch Proportions during Selection
We monitored egg hatch proportions for the first eight generations as an indication of the selection response. Overall hatch proportions during the first generation of selection were intermediate and varied considerably between individual females, but from the second generation of selection onwards they fell consistently below 4% ( Table 1). The wMel infection was lost in the selected line after the seventh generation of heat stress. No eggs hatched, and Wolbachia was not detected by qPCR in a sample of 30 individuals from the population. However, selection continued for a total of eight generations with a backup line.
Egg hatch proportions of females from the first generation of heat stress increased dramatically in subsequent gonotrophic cycles (Kruskal-Wallis: χ 2 = 26.616, df = 2, p < 0.001, Figure S2A). This restoration of compatibility over time indicates a recovery of Wolbachia within the female. However, we observed no such recovery in subsequent selection generations when initial hatch proportions were much lower ( Figure S2B), or during the first cytoplasmic incompatibility experiment, except for the wAlbB infection ( Figure 1). Egg hatch proportions of females from selection generations 2-7 were consistently low and did not change with female age according to a general linear model (F 1,65 = 0.629, p = 0.841, Figure S2B). We also tested if hatch proportions increased when heat stressed females were aged before mating with Wolbachia-infected males, but found no improvement with female age ( Figure S3).

Cytoplasmic Incompatibility and Restoration of Compatibility Following Selection
After five generations of selection, we tested the ability of wMel-HS males reared under different temperature conditions to induce cytoplasmic incompatibility ( Figure 1A,B). When larvae were reared at 26 • C, wMel and wAlbB males induced complete cytoplasmic incompatibility with uninfected females; no eggs hatched over two gonotrophic cycles. In contrast, some females that mated with wMel-HS males produced a low proportion of viable offspring, indicating incomplete cytoplasmic incompatibility induction by wMel-HS males. wMel and wMel-HS males induced partial cytoplasmic incompatibility when reared under heat stress, with weaker cytoplasmic incompatibility (and higher under all rearing conditions tested, with no eggs hatching in either gonotrophic cycle. We tested the ability of wMel-HS females reared under different temperature conditions to restore compatibility in crosses with infected males (Figure 1C,D). Hatch proportions of infected females reared at 26 °C were consistently high, and did not differ between the wMel, wMel-HS, and wAlbB lines (Kruskal-Wallis: χ 2 = 5.533, df = 2, p = 0.063). Egg hatch proportions declined when females were reared under heat stress conditions for the wMel (Mann-Whitney U: Z = 7.416, p < 0.001) and wMel-HS lines (Z = 7.720, p < 0.001) and to a lesser extent the wAlbB line (Z = 5.004, p < 0.001). The loss of compatibility was particularly severe at 26-37.5 °C where median hatch rates were below 4% for the wMel and wMel-HS lines. Egg hatch proportions did not differ between the wMel and wMel-HS lines at 26-36.5 °C (Z = 1.016, p = 0.308) or 26-37.5 °C (Z = 0.243, p = 0.810), indicating a lack of selection response. Egg hatch proportions did not increase for heat-stressed wMel (Z = 0.401, p = 0.689) or wMel-HS (Z = 0.422, p = 0.674) females in the second gonotrophic cycle, but did increase for wAlbB (Z = 2.131, p = 0.033), suggesting that Wolbachia strains can differ in their ability to recover from heat stress.  We tested the ability of wMel-HS females reared under different temperature conditions to restore compatibility in crosses with infected males (Figure 1C,D). Hatch proportions of infected females reared at 26 • C were consistently high, and did not differ between the wMel, wMel-HS, and wAlbB lines (Kruskal-Wallis: χ 2 = 5.533, df = 2, p = 0.063). Egg hatch proportions declined when females were reared under heat stress conditions for the wMel (Mann-Whitney U: Z = 7.416, p < 0.001) and wMel-HS lines (Z = 7.720, p < 0.001) and to a lesser extent the wAlbB line (Z = 5.004, p < 0.001). The loss of compatibility was particularly severe at 26-37.5 • C where median hatch rates were below 4% for the wMel and wMel-HS lines. Egg hatch proportions did not differ between the wMel and wMel-HS lines at 26-36.5 • C (Z = 1.016, p = 0.308) or 26-37.5 • C (Z = 0.243, p = 0.810), indicating a lack of selection response. Egg hatch proportions did not increase for heat-stressed wMel (Z = 0.401, p = 0.689) or wMel-HS (Z = 0.422, p = 0.674) females in the second gonotrophic cycle, but did increase for wAlbB (Z = 2.131, p = 0.033), suggesting that Wolbachia strains can differ in their ability to recover from heat stress.

Response of Laboratory and Field wMel Infections to Cyclical Heat Stress
We performed a second set of experiments to compare a wMel strain maintained in the laboratory for approximately 60 generations (wMel-lab) to a wMel strain that has been in the field for seven years (wMel-field) in their response to cyclical heat stress. We also tested the wMel-HS line, which had now been through eight generations of selection. When reared at 26 • C, both wMel-lab and wMel-field males induced complete cytoplasmic incompatibility, with no hatching eggs (Figure 2A). wMel-HS males reared at 26 • C induced nearly complete cytoplasmic incompatibility, though a single female produced one viable progeny. wMel-lab, wMel-field, and wMel-HS males did not differ significantly in their ability to induce cytoplasmic incompatibility when reared at 26-36 • C (Kruskal-Wallis: χ 2 = 5.534, df = 2, p = 0.063) or 26-37 • C (χ 2 = 4.858, df = 2, p = 0.088). All lines induced weaker cytoplasmic incompatibility when males were reared at 26-37 • C relative to 26-36 • C (Mann-Whitney U: Z = 5.072, p < 0.001).
Insects 2018, 9, x 7 of 12 both the first (C) and second (D) gonotrophic cycles. Black bars indicate median egg hatch proportions with 95% confidence intervals.

Response of Laboratory and Field wMel Infections to Cyclical Heat Stress
We performed a second set of experiments to compare a wMel strain maintained in the laboratory for approximately 60 generations (wMel-lab) to a wMel strain that has been in the field for seven years (wMel-field) in their response to cyclical heat stress. We also tested the wMel-HS line, which had now been through eight generations of selection. When reared at 26 °C, both wMel-lab and wMel-field males induced complete cytoplasmic incompatibility, with no hatching eggs ( Figure  2A). wMel-HS males reared at 26 °C induced nearly complete cytoplasmic incompatibility, though a single female produced one viable progeny. wMel-lab, wMel-field, and wMel-HS males did not differ significantly in their ability to induce cytoplasmic incompatibility when reared at 26-36 °C (Kruskal-Wallis: χ 2 = 5.534, df = 2, p = 0.063) or 26-37 °C (χ 2 = 4.858, df = 2, p = 0.088). All lines induced weaker cytoplasmic incompatibility when males were reared at 26-37 °C relative to 26-36 °C (Mann-Whitney U: Z = 5.072, p < 0.001).
We compared the ability of wMel-lab and wMel-field females to restore compatibility with infected males when reared under cyclical heat stress. wMel-lab and wMel-field females exhibited similar hatch proportions when reared at both 26-36 °C (Z = 0.759, p = 0.447) and 26-37 °C (Z = 0.135, p = 0.889, Figure 2B). In contrast, wMel-HS females had higher hatch proportions than wMel-lab females when they were reared at 26-36 °C (Z = 3.601, p < 0.001) though not at 26-37° C (Z = 0.095, p = 0.928). Perhaps the wMel-HS line has evolved an increase in ability to restore compatibility with infected males under 26-36 °C although not at the temperature cycle used for selection. Hatch rates were lower overall when females were reared at 26-37 °C compared to 26-36 °C, indicating a reduced ability of females to restore compatibility at higher temperatures (Z = 6.563, p < 0.001). Males reared at each temperature were crossed to uninfected females to test their ability to induce cytoplasmic incompatibility; (B) Females reared at each temperature were crossed to infected males reared at 26 °C to test their ability to restore compatibility. Black bars indicate median egg hatch proportions with 95% confidence intervals.

Wolbachia Density
We tested if Wolbachia density had changed in response to selection. After five generations of selection, we estimated Wolbachia density in whole wMel and wMel-HS adults when reared at a constant 26 °C or under two cyclical temperature conditions. These adults were taken from the same Figure 2. The ability of wMel to induce cytoplasmic incompatibility and restore compatibility under heat stress following field deployment and selection. Males (A) and females (B) from the wMel-lab, wMel-field, and wMel-HS (following eight generations of selection) lines were reared at a constant 26 • C, or held in incubators that cycled diurnally between 26 • C and 36 • C or 26 • C and 37 • C; (A) Males reared at each temperature were crossed to uninfected females to test their ability to induce cytoplasmic incompatibility; (B) Females reared at each temperature were crossed to infected males reared at 26 • C to test their ability to restore compatibility. Black bars indicate median egg hatch proportions with 95% confidence intervals.
We compared the ability of wMel-lab and wMel-field females to restore compatibility with infected males when reared under cyclical heat stress. wMel-lab and wMel-field females exhibited similar hatch proportions when reared at both 26-36 • C (Z = 0.759, p = 0.447) and 26-37 • C (Z = 0.135, p = 0.889, Figure 2B). In contrast, wMel-HS females had higher hatch proportions than wMel-lab females when they were reared at 26-36 • C (Z = 3.601, p < 0.001) though not at 26-37 • C (Z = 0.095, p = 0.928). Perhaps the wMel-HS line has evolved an increase in ability to restore compatibility with infected males under Insects 2018, 9,78 8 of 13 26-36 • C although not at the temperature cycle used for selection. Hatch rates were lower overall when females were reared at 26-37 • C compared to 26-36 • C, indicating a reduced ability of females to restore compatibility at higher temperatures (Z = 6.563, p < 0.001).

Wolbachia Density
We tested if Wolbachia density had changed in response to selection. After five generations of selection, we estimated Wolbachia density in whole wMel and wMel-HS adults when reared at a constant 26 • C or under two cyclical temperature conditions. These adults were taken from the same containers used for the first cytoplasmic incompatibility experiment. Wolbachia density was greatly reduced under heat stress in both wMel (Mann-Whitney U: Z = 8.899, p < 0.001) and wMel-HS (Z = 8.885, p < 0.001) lines ( Figure 3). wMel-HS males had lower Wolbachia densities than wMel males at 26 • C (Z = 3.544, p < 0.001), 26-36.5 • C (Z = 5.396, p < 0.001), and 26-37.5 • C (Z = 3.746, p < 0.001). For females, Wolbachia density did not differ between the wMel and wMel-HS lines (all p > 0.05) except for when they were reared at 26-36.5 • C, where the wMel line had a higher density than the wMel-HS line (Z = 4.720, p < 0.001).  We measured the Wolbachia density of wMel-lab, wMel-field, and the wMel-HS line after eight generations of selection. Wolbachia density was greatly reduced at higher rearing temperatures for all strains (Kruskal-Wallis: df = 2, all p < 0.001, Figure 4). Wolbachia density in the wMel-field and wMellab lines did not differ at each temperature and for each sex (Mann-Whitney U: all p > 0.05), indicating that the infection density of wMel has remained stable following field releases. In contrast, Wolbachia density was greatly reduced in wMel-HS females (Mann-Whitney U: Z = 5.396, p < 0.001) and males (Z = 5.234, p < 0.001) at 26 °C relative to wMel-lab, with wMel-HS females and males having 16% and 25% the density of wMel-lab females and males, respectively. wMel-HS adults had a higher density than wMel-lab adults when reared at 26-36 °C (Z = 4.566, p < 0.001) but a lower density when reared at 26-37 °C (Z = 2.839, p = 0.005); any effects of selection on density therefore appear to be inconsistent across conditions. We measured the Wolbachia density of wMel-lab, wMel-field, and the wMel-HS line after eight generations of selection. Wolbachia density was greatly reduced at higher rearing temperatures for all strains (Kruskal-Wallis: df = 2, all p < 0.001, Figure 4). Wolbachia density in the wMel-field and wMel-lab lines did not differ at each temperature and for each sex (Mann-Whitney U: all p > 0.05), indicating that the infection density of wMel has remained stable following field releases. In contrast, Wolbachia density was greatly reduced in wMel-HS females (Mann-Whitney U: Z = 5.396, p < 0.001) and males (Z = 5.234, p < 0.001) at 26 • C relative to wMel-lab, with wMel-HS females and males having 16% and 25% the density of wMel-lab females and males, respectively. wMel-HS adults had a higher density than wMel-lab adults when reared at 26-36 • C (Z = 4.566, p < 0.001) but a lower density when reared at 26-37 • C (Z = 2.839, p = 0.005); any effects of selection on density therefore appear to be inconsistent across conditions. density was greatly reduced in wMel-HS females (Mann-Whitney U: Z = 5.396, p < 0.001) and males (Z = 5.234, p < 0.001) at 26 °C relative to wMel-lab, with wMel-HS females and males having 16% and 25% the density of wMel-lab females and males, respectively. wMel-HS adults had a higher density than wMel-lab adults when reared at 26-36 °C (Z = 4.566, p < 0.001) but a lower density when reared at 26-37 °C (Z = 2.839, p = 0.005); any effects of selection on density therefore appear to be inconsistent across conditions.

Discussion
In this study, we attempted to generate a heat-resistant strain of wMel in Ae. aegypti that could serve as an alternative Wolbachia variant for disease control programs. In response to intense selection, wMel did not consistently improve its thermal tolerance and only exhibited a reduced density at 26 • C compared to an unselected wMel strain. We also show that wMel infections from the laboratory and field did not differ in their response to heat stress or in their phenotypic effects under standard laboratory conditions, despite the infection having been maintained in field populations for seven years. These findings suggest that the wMel infection is unlikely to change much in its response to heat stress after field deployment. On the other hand, our comparisons between the lab strain and field strain indicate that the infection will likely remain stable in its ability to induce cytoplasmic incompatibility under non-stressful conditions for many years following field releases.
Given that Wolbachia strains can differ substantially in their response to cyclical heat stress [28] and adapt rapidly to changing environments [35,37], the limited response of wMel to selection for increased heat tolerance in the laboratory despite intense selection is perhaps surprising. It is possible that trait-relevant genetic variation within both the wMel Wolbachia genome and nuclear Aedes genome is limited. Drosophila data suggest a limited ability of populations to adapt to high temperature stress, which may be related to a lack of genetic variation generally or genetic interactions with other traits [45]. Perhaps adaptation to heat stress might occur after a much longer period of selection, particularly given that females from the selected line seemed to show an increase in compatibility with infected non-stressed males after eight selection generations when the females were reared at 26-36 • C.
The comparison of field and lab strains showed no change in responses to heat stress despite lines being separated by 60 generations. In the field, the strength of selection for maintaining a high density of Wolbachia is unclear given that this will depend on exposure to high temperatures in undefined breeding sites. Field experiments suggest that when mosquitoes breed in Cairns during the wet season, Wolbachia will reduce in density sufficiently to trigger a loss of compatibility with infected males in breeding containers placed in full or partial sunlight (Ross, unpublished). However, it is not known what proportion of a mosquito population will be exposed to sufficiently high temperatures. Moreover, any costs of increased heat resistance in terms of a lower density of Wolbachia at 26 • C could reduce selection pressures for resistance in the case where females suffer reduced egg hatch when mated to infected males.
Our results for the field strain comparison are consistent with previous studies showing that the wMel infection has remained stable in its effects after field deployment [13,14], but contrast with the several examples of Wolbachia infections adapting rapidly to changes in environmental conditions in other insects. There are concerns that Wolbachia strains in Aedes mosquitoes could evolve and lose effectiveness after field deployment [46]. We suspect that such evolution is unlikely in the short-term, at least for the wMel infection, given that its density and ability to induce cytoplasmic incompatibility have remained stable for seven years in both the laboratory and the field.
Despite maintaining susceptibility to heat stress, the wMel infection has established and persisted at a high frequency in multiple locations in Cairns, Australia [2,13], but this does not ensure the success of releases in other areas where maximum daily temperatures during the hottest times of year are higher than in release zones in Australia. Since weather reports are not necessarily good indicators of microclimates in mosquito larval habitats [47], careful monitoring will be required to determine if heat stress will be a concern for wMel release programs in other areas. The deployment of alternative strains like wAlbB that are more thermally robust may be required in these areas for efficient population transformation and viral suppression.
Supplementary Materials: The following are available online at http://www.mdpi.com/2075-4450/9/3/78/s1, Figure S1: Temperatures experienced in larval rearing trays during the (A) first and (B) second cytoplasmic incompatibility experiments on Ae. aegypti, Figure S2: Proportion of eggs hatching from crosses between wMel-infected Ae. aegypti females reared under heat stress and wMel-infected males reared at 26 • C during the selection experiment, Figure S3: Hatch proportions of eggs from wMel-infected female Ae. aegypti reared under cyclical heat stress and held at 26 • C for 1, 4 or 7 days before being crossed to wMel-infected males reared at 26 • C, Figure S4: Wolbachia density of wMel, wMelPop and wAlbB strains when Ae. aegypti are reared under heat stress and provided with different levels of larval nutrition.