top of page

Mosquito immunity, physiology, and the pathogen interaction


Over the past three decades, we have gained substantial insights into the mosquito proteins and cellular machinery mediating mosquito-pathogen interactions. Yet, we are just beginning to understand the complexity underlying mosquito-pathogen interactions. Mosquito infection is a dynamic phenotype, which is dependent upon both the specific mosquito-pathogen pairing, as well as in variation in key environmental factors. This is especially relevant for understanding how specific pathogens emerge, environmental conditions that favor emergence, and the biological constraints on the distributions of emerging mosquito-borne diseases. For example, there is overwhelming evidence that temperature markedly affects diverse aspects of mosquito physiology, life history, and pathogen replication within the mosquito because mosquitoes are small, cold-blooded organisms. Yet, the extent to which temperature shapes transmission directly, through effects on pathogen biology, or indirectly, through effects on mosquito immunity and physiology, remain largely unexplored.


Aedes aegypti - Zika virus


Anopheles stephensi - human malaria


Culex quinquefasciatus - West Nile virus

Aedes and Anopheles photos are credited to iStock (Getty images) and, respectively. Culex photo is from the Mississippi Entomological Museum. Human malaria gametocyte images taken from Coatney GR, Collins WE, Warren M, Contacos PG. The Primate Malarias. Bethesda: U.S. Department of Health, Education and Welfare; 1971. RNA-seq illustration taken from Oxford Gene Technology. West Nile virus image taken from The Native Antigen site.

We currently have two ongoing projects that utilize RNA sequencing and bioinformatic analysis to identify panels of differentially expressed genes that are important in the physiological response of mosquitoes to temperature, to infection, and how temperature alters mosquito responses to infection in the yellow fever mosquito (Aedes aegypti) - Zika virus and the Anopheles stephensi (Indian malaria mosquito) - human malaria (Plasmodium falciparum) systems. We also are initiating a new project exploring the effects of temperature variation on West Nile virus evolution in North American Culex spp.

Environmental drivers of mosquito life history and disease transmission


There are several key knowledge gaps that affect our ability to predict and, ultimately, mitigate the factors influencing vector-borne disease transmission. This is particularly important in systems where we lack fundamental knowledge on the relationships between key environmental variables and transmission. A large component of the research conducted in our group focuses on the effects of abiotic and biotic variation on mosquito life history and overall transmission.

A melanized sephadex bead

expression of DEFCEC, and NOS

with heat-killed E. coli

1.) Temperature effects on Zika virus transmission and control


Mosquito-borne viruses are an emerging threat impacting human health and well-being. Diseases such as Zika, dengue, and chikungunya, which were once considered tropical and sub-tropical diseases, are now threatening temperate regions of the world due to climate change, globalization, and increasing urbanization. It is estimated that 3.9 billion people in 120 countries around the globe are at risk of contracting an arboviral disease. In spite of increasing efforts to develop vaccines and therapeutics, vector control is still the only way to mitigate the disease spread. Therefore, it is crucial to try to predict how these viruses might spread seasonally, geographically, and with climate change. We used empirical and mathematical modeling approaches to predict how Zika virus transmission changes with temperature.


Zika virus (ZIKV, dark blue) transmission has a non-linear relationship with temperature, with a similar temperature optimum where transmission is maximized and temperature maximum where temperature is minimized as other mosquito-borne viruses like dengue (DENV, light blue). Thus, as environmental temperatures warm and seasons extend with climate change and urbanization, environmental conditions at higher latitudes will become more favorable for Zika virus transmission. However, because the minimum temperature for Zika virus transmission is 5 C higher than for dengue virus,  the expansion of transmission risk at higher latitudes will be less dramatic than for dengue virus (Tesla et al. 2018 Temperature drives Zika virus transmission: evidence from empirical and mathematical models. Proceedings of the Royal Society Series B).


Using the temperature-dependent model developed above to predict the effects of future climate warming on the global environmental suitability for Zika virus transmission, we find a substantial increase in the global population that could live in areas suitable for this virus.  Under modest (RCP 4.5) to severe (RCP 8.5) warming scenarios and model predictions for population change in at risk areas, 2.5-2.7 billion more people are expected to be at risk for contracting Zika virus in the future.

Environment, Body Condition, & Vectorial Capacity - Highlight # 2


Vectorial capacity is a measure of the transmission potential of a mosquito population, is defined by the following equation, and is comprised of both mosquito and parasite traits:




This figure highlights predicted changes in environmental suitability for North, Central, and South America by 2050 under modest and severe warming scenarios (Ryan et al. 2020 Warming temperatures could expose 1.3 billion new people to Zika virus risk by 2050. Global Change Biology).

Whether or not temperature affects the potential for disease control is an important applied question for designing public health campaigns (either via vector control, reduction in host biting rate, vaccination, or drug administration). Work from our group demonstrates that the effectiveness of most parameters that are sensitive to disease control measures depend on temperature, even those not thought to be directly sensitive to temperature variation (e.g., human vaccination rate). In particular, dynamical, mechanistic models predict that the human vaccination rate required to control Zika epidemics varies strongly with mean temperature. Thus, the efficacy of different intervention strategies and the population coverage required will likely need to be higher in environments that are highly suitable for transmission, which include areas of the world or times of season when temperature is near the thermal optimum for transmission of a given disease (e.g., Zika virus). 

Hugh Sturrock


Preliminary work suggests that changes in mosquito body condition can have profound effects on vectorial capacity through changes in both mosquito and parasite traits. Effects on mosquito traits are the following:

The human final epidemic size is sensitive to the annual mean (oscillation) temperature, Tm, and several important control-related parameters: (a) the seasonal divergence of the annual temperature from the mean, Ta; (b) the human recovery rate, γh; (c) the human vaccination rate, δh; (d) the scaling factor of vector biting rate, cbv; (e) the probability of transmission from the mosquito to the human, βvh; (f ) the scaling factor of the probability of transmission from the human to the mosquito, cβhv; (g) the scaling factor of the vector mortality rate, cls; (h) the vector carrying capacity, κv; and (i) the egg survival probability scaling factor, cνv. The annual mean (oscillation) temperature varies along the x-axis, the seasonal divergence of the annual temperature from the mean and the other parameters that are sensitive to control measures vary along the y-axes, and the color scale indicates the total infectious humans. Apart from (a), where the temperature amplitude (Ta) is varying, the amplitude is set at 10°C for the other plots. Plot (a) shows that there can be large epidemics even when mean temperatures are low if the seasonal variation (the amplitude) is high enough, as would be found in subtropical and temperate regions (Ngonghala et al. 2021 Effects of changes in temperature on Zika virus dynamics and control. Journal of the Royal Society Interface).

Effects on parasite associated traits are the following:


2.) Spatial heterogeneity in abiotic and biotic factors and implications for  context-

     dependent interactions, transmission, and control.


The global distribution of mosquito species and the pathogens they transmit are strongly determined by climatic factors such as temperature and rainfall. However, fine-scale spatial variation in vector-borne disease dynamics can arise due to limited mosquito dispersal ability and the patchiness of available, high quality habitat. The microclimates mosquitoes experience, the availability and quality of habitat, and access to vertebrate hosts can differ dramatically over short distances due to variation in land cover and land use. This fine-scale heterogeneity can impact mosquito population dynamics, human exposure to biting mosquitoes, and disease transmission. Thus, understanding these fine-scale dynamics is essential for accurate forecasting of disease risk and evaluating the effects of climate and land use change on vector-borne disease transmission.  Our lab examines the effect of variation in abiotic (e.g., microclimate) and biotic factors (e.g., habitat availability, inter- and intra-specific competition) on mosquito population dynamics, genetic structure, and arbovirus transmission potential of Ae. aegypti (Yellow fever mosquito) and Ae. albopictus (Asian tiger mosquito).















Current work in this theme is ongoing with Ae. albopictus (Asian tiger mosquito) around Atlanta (Georgia), Athens (Georgia), and Long Island (New York), with Ae. aegypti (Yellow Fever mosquito) on St. Kitts & Nevis, and with An. stephensi (Asian urban malaria vector) in Surat and Ahmedabad, India. 


Example breeding site locations for the Asian tiger mosquito (Ae. albopictus) in Georgia. Pictures taken by Courtney Murdock

3.) Resolving uncertainty and refining transmission models of malaria


The deadliest organism on the planet is the Anopheles mosquito, the insect that transmits malaria to humans. Human malaria is the leading killer among infectious diseases, resulting in approximately 216 million cases and 500,000 deaths, primarily in children under the age of 5. Dramatic reductions in disease burdens of human malaria in the last 20 years has led to ambitious calls for eradicating the disease by 2030. However, eradication hinges on our ability to eliminate transmission 1) from both symptomatic and asymptomatic hosts, 2) across highly heterogeneous landscapes created by biotic and abiotic factors, and 3) with the emergence of multi-drug resistance to the last remaining anti-malarials, artemisinin-combination therapies. Despite growing research efforts to develop new therapeutics and vaccines, mitigating malaria transmission still largely depends on conventional mosquito control methods (e.g. bed nets, indoor residual spraying). Developing tools that will allow us to successfully predict outbreaks and efficiently target current and future interventions to specific times and locations will aid effective mosquito and disease control.


Current work revolves around experimentally validating common assumptions made by mathematical models that predict transmission, identifying additional sources of ecological variation these models should incorporate, and building new models to improve prediction.  This includes assessing how transmission models perform when key mosquito life history traits (e.g. daily per capita mortality rate, egg production, and biting rate) change across the lifespan of the mosquito vector, in thermally variable environments, across different species of Anopheles mosquitoes, with parasite infection, and in response to interacting environmental variables. Additionally, we are interested in exploring the impact of individual variation in these life history traits and the presence of life history trade-offs across these traits on malaria transmission.


The predicted temperature optimum that maximizes malaria transmission and the temperature maximum that minimizes malaria transmission differ when mosquito life history traits (e.g. lifespan, lifetime egg production, and daily biting rate) are assumed to vary across realistic environments that incorporate daily temperature ranges (dtr) of 9 C and 12 C relative to environments that are set to constant average temperature (dtr 0 C). Further, overall malaria transmission risk is predicted to differ across the distribution of Anopheles stephensi (the Indian malaria mosquito) depending on whether models estimate key parameters across multiple mosquito species (J estimated, peach color) or a single species (An. stephensi; MSP models, red and purple). Less differences in overall malaria transmission risk were noted between models that incorporate trait data estimated across the lifespan of the mosquito (MSP lifetime, purple) relative to models that estimate these traits (MSP estimated, red). Figures developed by Kerri Miazgowicz and Dr. Sadie Ryan (University of Florida).

Assessing the effectiveness of novel mosquito control tools

Many vector-borne diseases currently lack effective therapeutics and vaccines (e.g. the arboviruses), and global malaria elimination efforts are being threatened by the evolution of mosquito and parasite resistance to current interventions. Thus, there is a pressing need to develop novel intervention strategies to control these diseases. Engineered traits that render mosquitoes ineffective at transmitting parasites and pathogens (e.g. life shortening, parasite or pathogen blocking, upregulated immunity) are a promising new avenue of research in vector-borne disease elimination efforts. The success of many of these technologies relies on mosquito release in the field and subsequent invasion of transgenic / transinfected traits in wildtype mosquito populations. These technologies often fail in the field either due to a poor understanding of how environmental variation shapes transgenic / transinfected traits or what female mosquitoes find sexy in their male mates (i.e. mating biology).

1.) Exploring implications of mosquito love songs and mate choice on fitness

This project is a series of exploratory studies investigating whether female mosquitoes utilize acoustic signals in mate choice and if these signals serve as reliable indicators of male fitness and offspring viability. We, in collaboration with Drs. Laura Cator (Imperial College) and Laura Harrington (Cornell University), are defining the relationship between newly identified acoustic courtship signals and the fitness of both males and their offspring in the Aedes aegypti - dengue virus system. At the University of Georgia, we are investigating the effects of female mate choice on offspring immune performance, a key fitness and transmission metric. We are interested in identifying if harmonic convergence is a reliable cue for male offspring immune performance and whether life history trade-offs exist between immunity and reproductive effort. The potential findings of this collaborative research will serve as a basis for a research program designed to facilitate application of acoustics in the assessment and improvement of proposed release lines.

Edited Image 2015-1-19-10:14:9
Edited Image 2015-1-19-10:12:54
Edited Image 2015-1-19-10:11:5
humoral melanization 
bacterial killing
ability to transmit dengue virus

The buzz of a flying female mosquito acts as a mating signal, attracting males. The important behavioral component of the buzz is the fundamental frequency of the mosquito wing beat. This fundamental frequency typically occurs between 300-600 Hz depending on the mosquito species. For Aedes aegypti, the fundamental frequency for females is approximately 400 Hz and 600 Hz for males. Prior to mating, Ae. aegypti males and females modulate their flight tones when brought within a few centimeters of each other. Further, this modulation does not match the fundamental wing beat frequency of the male or female, but represents a shared harmonic of around 1200 Hz (A, Cator et al. 2009. Harmonic convergence in the love songs of the dengue vector mosquito. Science).  Pairs that harmonically converge are more likely to successfully mate (B, 1). Additionally, female mosquitoes also display a variety of rejection behaviors towards males that are deemed unsuitable as mates. These include actively kicking (B, 2) or holding (B, 3) the male away from her as he attempts to mate (Cator et al. 2011. The harmonic convergence of fathers predicts the mating success of sons in Aedes aegypti. Animal Behaviour). We are currently investigating whether or not offspring from parental pairs that harmonically converge differ in humoral (e.g. melanization neutrally charged sephadex beads) and cellular (e.g. ability to kill Escherichia coli) immune responses, as well as their susceptibility and resistance to dengue-2 virus (C, photos are provided by Courtney Murdock, and the dengue virus image was taken from the Protein Data Bank in Europe website).

2.) The effect of environmental variation on malaria control strategies

Artemisinin is the last line of anti-malarials against human malaria (Plasmodium falciparum) where resistance is characterized by delayed recovery times and increased number of parasites in patients infected with resistant strains. Although parasite strains vary widely in their ability to produce transmission stages of the parasite, numerous laboratory and field studies investigating the relationship between parasite numbers and probability of infection in the mosquito vector have collectively led to the description of a positive, albeit loose, log-linear relationship. This, in turn, has led to the general view that hosts with higher numbers of parasites contribute more to transmission, and the concern that the elevated parasite numbers observed in patients infected with artemisinin-resistant parasites could be contributing to the rapid proliferation of artemisinin-resistance in response to intense drug pressure. As a result, the WHO recommends treating patients with primaquine, the only known gametocide to date, in addition to artemisinin. 


However, the significant amount of variation around the positive relationship between parasite density and mosquito transmission suggests 1) low-density carriers introduce significant heterogeneity in transmission, 2) the relationship reaches an eventual asymptote where higher parasite numbers offer no apparent, additional benefits, and 3) there are still many unexplained factors influencing successful transmission of parasites to mosquitoes. In addition to biotic factors, such as variation in host immuno-physiology that allow for certain hosts to be highly infectious regardless of parasite densities, abiotic factors could influence this relationship. We are interested in exploring the relative effects of environmental temperature on this relationship, as well as the implications of environmental variation for the spread of artemisinin resistant parasites and the effectiveness of tools that target transmission stages of the parasite (e.g. primaquine and transmission blocking vaccines).


To map fine-scale variation in temperature suitability for arbovirus transmission, we use a temperature-dependent vectorial capacity (VC(T)) expression that uses a combination of data generated from field studies, laboratory experiments, and mechanistic modeling. Orange boxes represent input data sets that are generated from field studies that use a combination of data logging, mosquito larval surveys, and mosquito adult surveys to characterize the effect of land use and land cover on fine-scale variation in mosquito microclimate (temperature and relative humidity), the abundance of mosquito habitat, and relative adult abundance. Green boxes represent both empirical models that utilize statistical relationships that exist between microclimate and mosquito relative abundance (mosquito abundance model) and land cover / use and macroclimate with mosquito relevant microclimate (microclimate model), and mechanistic models that utilize observed relationships between temperature and key mosquito and pathogen traits relevant for transmission (temperature-trait curves). Blue boxes represent model predictions. Gridded spatial predictions from the microclimate model, the mosquito abundance model, as well as temperature-trait relationships characterized in the lab for other key traits can be used estimate variation in thermal suitability for arbovirus transmission at scales appropriate for the transmission process and for implementing disease control (Wimberly et al. 2021 Land cover affects microclimate and temperature suitability for arbovirus transmission in an urban landscape. PLoS Neglected Tropical Diseases; Evans et al. 2019 Microclimate and larval habitat density predict Ae. albopictus abundance in urban areas. American Journal of Tropical Medicine and Hygiene; Tesla et al. 2018 Temperature drives Zika virus transmission: evidence from empirical and mathematical models. Proceedings of the Royal Society London Series B).

bottom of page