Know your enemy: Environmental variation significantly affects mosquito biology and malaria transmission risk

Open Access
Shapiro, Lillian Lee
Graduate Program:
Doctor of Philosophy
Document Type:
Date of Defense:
August 31, 2016
Committee Members:
  • Matthew B. Thomas, Dissertation Advisor
  • Matthew B. Thomas, Committee Chair
  • Andrew F. Read, Committee Member
  • Jason L. Rasgon, Committee Member
  • James H. Marden, Outside Member
  • malaria
  • mosquito
  • climate change
  • disease ecology
  • Anopheles
  • Plasmodium
The eradication of malaria remains one of the most substantial challenges facing the scientific community in the 21st century. Each year, malaria kills more than 500,000 people (mostly children under the age of five), while another 3.3 billion remain at risk. At the core of many complex issues surrounding malaria control and eradication is the biology of the mosquito vector, which is quite sensitive to shifts in a number of environmental variables, such as temperature, humidity, nutrition, and land use changes. An integrative perspective of the variation in mosquito biology and ecology is crucial to our understanding of malaria transmission, especially in the context of a changing climate. Here, we examine how environmental variability across the mosquito life cycle affects mosquito life history, malaria parasite development, and ultimate disease transmission risk. We found that variability in larval food quantity results in considerable changes in both larval and adult mosquito traits. Reduction of larval nutrients greatly extends the time until adult emergence, and habitats with less food produce about 25% fewer adult females than habitats with optimal amounts of food. Females that emerge from nutritionally-stressed habitats have higher rates of mortality, are less likely to mate and lay eggs, and exhibit longer intervals between blood meals (which dictates the number of probable infectious bites a mosquito may transmit in her lifetime). We also observed an effect of larval food quantity on the intensity of infection when mosquitoes were exposed to the rodent malaria Plasmodium yoelii, but there was no change in overall parasite prevalence. However, the changes in life history traits alone led to a 70% reduction in vectorial capacity in females from low food larval environments compared to those from optimal food environments. When we infected females from similar dietary regimes with the human malaria parasite Plasmodium falciparum, we observed that parasite prevalence in adults from reduced food environments is generally lower, and that parasite development in low-food females is up to two days slower than in females from high-food larval environments. We assessed malaria transmission potential using an epidemiological model that focuses on mosquito mortality and parasite development over time. Our results show that, over a 30-day period, transmission potential is increased by 330% in a population of females emerging from high larval food environments compared to an equivalently sized population of females that experienced nutritional stress as larvae. We observed more complex effects on transmission potential when we integrated two types of environmental variation at the larval stage (food quantity and temperature). The optimum temperature for larval and adult mosquito fitness and development of the human malaria parasite Plasmodium falciparum within the mosquito vector is generally considered to be approximately 27ºC. However, we found that adult females from hot (32°C), high larval food environments had the lowest mortality rates when transferred back to the “optimum” 27ºC, while the highest biting rate was observed in adult females from cool (22°C), high larval food environments. These results suggest that there is no “optimum” larval mosquito habitat, but rather, that different habitats optimize different traits. We also found that, despite the common assumption that larger mosquitoes make better malaria vectors, vectorial capacity is only weakly correlated with size, and is far more heavily impacted by the differences in larval environment that generate the observed variation in female size. Finally, we exposed mosquitoes infected with P. falciparum to seven temperatures ranging from 18°C to 34°C. We found that the transmission potential of mosquitoes was highest in mid-range environments (24-27°C), while transmission potential decreases steeply on both sides of this optimum range, and no infectious mosquitoes were observed at 18ºC. Our results challenge previous predictions of parasite development thresholds at both the cool and warm ends of the temperature spectrum. Our findings also suggest that malaria transmission in some endemic areas may decrease as temperatures warm, which contrast many previous predictions that a warming climate will enhance malaria transmission in endemic areas. However, the geographic area that is optimal for malaria transmission may expand, as previously cooler temperatures warm towards a thermal optimum. Overall, our results demonstrate that a more nuanced understanding of how environmental variation shapes mosquito biology and ecology is crucial in order to make optimum use of current vector control resources and to develop innovative, more sustainable control strategies for the future.