The effect of temperature on Anopheles mosquito population dynamics and the potential for malaria transmission

Abstract

The parasites that cause malaria depend on Anopheles mosquitoes for transmission; because of this, mosquito population dynamics are a key determinant of malaria risk. Development and survival rates of both the Anopheles mosquitoes and the Plasmodium parasites that cause malaria depend on temperature, making this a potential driver of mosquito population dynamics and malaria transmission. We developed a temperature-dependent, stage-structured delayed differential equation model to better understand how climate determines risk. Including the full mosquito life cycle in the model reveals that the mosquito population abundance is more sensitive to temperature than previously thought because it is strongly influenced by the dynamics of the juvenile mosquito stages whose vital rates are also temperature-dependent. Additionally, the model predicts a peak in abundance of mosquitoes old enough to vector malaria at more accurate temperatures than previous models. Our results point to the importance of incorporating detailed vector biology into models for predicting the risk for vector borne diseases.

Figure 1. Diagram of the model setup.

Each stage experiences temperature-dependent, stage-specific mortality, ( or ). Recruitment into a stage at time is given by and is also dependent on temperature. Density-dependent mortality is only experienced in the larval stage, .

Figure 2. Larval and adult equilibrium abundances.

(A) The larval equilibrium abundances across temperatures with exponential density-dependence. (B) The adult equilibrium abundances (solid line, left axis) and daily survival (dashed line, right axis) across temperatures. The gray points and bars in both panels are the stable and cyclic abundances, respectively. The solid line connecting the points is the average abundance across temperature. Notice that the y-axes have different scales.

Figure 3. Age-specific adult abundance and adult recruitment across temperature.

(A) The age-specific abundance of a single cohort of adult mosquitoes for each temperature. High abundance is in dark blue decreasing to zero in white. (B) Recruitment (the mean abundance of new recruits) into the adult stage over the temperature range.

Figure 4. Extrinsic incubation period curves.

Temperature-dependent extrinsic incubation period in days; the solid line is the Detinova prediction curve and the dashed line is the Paaijmans curve.

Figure 5. Potential for infectious mosquitoes.

The abundance of mosquitoes old enough to be potentially infectious across temperatures. In both graphs the solid line represents the predictions made using our model in combination with either the Detinova (A) or Paaijmans (B) EIP prediction curve. The gray points and bars in both panels are the stable and cyclic abundances predicted by our model, respectively. The dashed lines represent the predictions made using the classic model assumptions of a constant vector population, in combination either the Detinova (A) or Paaijmans (B) EIP prediction curve. The red points, which correspond to the right y-axis are the observed entomological inoculation rates from 14 countries in Africa, compiled by .

Figure 6. Sensitivity rank across temperature.

The sensitivity of larval abundance (A), adult abundance (B) and adult recruitment (C) to changes in the parameters across temperatures ranked from highest to lowest sensitivity. Red indicates greatest sensitivity to change, followed by orange, yellow and white. The x-axis is temperature from 17–33°C, and the y-axis is the parameter.