14.2 Transmission

The second topic we must tackle is blood feeding, which is an interaction between mosquitoes and humans. It is an asymmetric relationship – mosquitoes search for blood hosts, select a host, and blood feed. Humans, for their part, attract mosquitoes from a distance, move around, and spend time in places when mosquitoes are biting. Humans can wear protective clothing (or not), use bed nets (or not), or do other things that make them more or less available to humans. Despite all this, humans are often unaware that they have been bitten.

Transmission occurs during blood feeding, and models of blood feeding should be able to take all this heterogeneity into account. If the models do a proper accounting, then the total number of human blood meals taken by mosquitoes would equal the number of blood meals received by humans. In doing so, we find no inspiration from Macdonald, whose description of human blood feeding was simple and phenomenological: a single parameter described the human blood feeding rates. After Garrett-Jones described the human blood index, drawing on decades of work, the one parameter was split into an overall blood feeding rate (\(f\)) and a human fraction (\(q\)). The question left unaddressed by Macdonald was how these rates vary by context, and the consequences for exposure. To do this, we reformulated the algorithm describing blood feeding [15].

Over the past two decades, several papers have drawn attention to the way blood feeding behaviors are or ought to be constrained by the availability of vertebrate hosts. It may be fine to assume that the density of vertebrate hosts doesn’t change, but something should change when a large fraction of people are using bednets. Even with static parameters, we should think through the limiting cases: if there are no vertebrate hosts, then there blood feeding should not occur (i.e., \(f=0\)); if there are no human hosts, then there should be no human blood meals (\(q=0\)); and if there are no alternatives to humans, all blood meals should be on humans (\(q=1\)).

The concepts we devised for blood feeding must, therefore, integrate the notion of frailty with the process of mosquito search. On the one hand, the mosquitoes should blood feed at a slower rate if hosts are unavailable. On the other hand, human biting should become heterogeneous. To arrive at an adequate description, we need to formalize this notion of host availability. The logic is that mosquitoes search for humans. Differences among humans in their attractiveness are represented by a search weight. Mosquito search in a place depends on the amount of time spent by humans, but also by daily mosquito activity patterns; from these, we develop a notion of time at risk that characterize the way human activities expose them to mosquitoes. The mosquitoes add up all the time at risk spent by all the humans, which gives a measure of their availability. Availability describes humans as well as other vertebrate hosts, which are modified by mosquito preferences. The overall feeding rates and the human fraction are computed from availability using functional responses.

To complete the picture, we consider how the expected rate of exposure could have a distribution in the population, which we call environmental heterogeneity.

References

15.
Wu SL, Henry JM, Citron DT, Ssebuliba DM, Nsumba JN, C HMS, et al. Spatial dynamics of malaria transmission. PLOS Computational Biology. 2023;19: e1010684. doi:10.1371/journal.pcbi.1010684