A model of Anopheles gambiae populations dynamics coupled with Plasmodium falciparum transmission dynamics is extended to include mechanisms of larval flushing which are known to occur. Flushing dynamics are modeled using a simulation that incorporates seasonal, autocorrelated, and random components based on 30 years of rainfall data for the Kakamega District of the western Kenya highlands. The model demonstrates that flushing phenomena can account for differences between regions with the same annual larval habitat pattern, changing the World Health Organization endemicity classification from either hyperendemic or holoendemic to hypoendemic disease patterns. Mesoendemic patterns of infection occur at the boundary of the holoendemic to hypoendemic transition. For some levels of flushing the entomological inoculation rate drops to an insignificant amount and disease disappears, while the annual indoor resting density remains well above zero. In these scenarios, the disease is hypoendemic, yet the model shows that outbreaks can occur when disease is introduced at particular time points.
Keywords: Anopheles gambiae; Flushing; Malaria; Mathematical model; Plasmodium falciparum.
© 2022 The Author(s).