Fig. 1: Metapopulation model for RVF virus transmission in the Comoros archipelago.
In order to quantify the drivers of Rift Valley fever in the Comoros archipelago, we developed a metapopulation model describing RVF virus infection of livestock (cattle, sheep and goats) in the Comoros archipelago. a We modelled the explicit movement of livestock (solid black arrows) between the four islands in the Comoros archipelago: Grande Comore (red), Mohéli (blue), Anjouan (green) and Mayotte (purple). b Within-island viral transmission was modelled as an age-stratified Susceptible-Exposed-Infected-Recovered (SEIR) model. The shown schematic illustrates the transfer of animals (arrows) between four infection states–susceptible, S, exposed, E, infected, I, and recovered, R–and 10 age groups, a = 1, ..., 10, (boxes) on each island. The number of individuals in each compartment was updated in discrete time, t, with a single time step equal to one epidemiological week. Animals were born into the youngest age group at a time-varying rate νt and were removed from the system due to death at an age-dependent proportion μa. A fixed proportion, δ, of individuals were aged at each time step. Susceptible animals were exposed to the disease at a time-varying proportion, λt, which was dependent on the mean Normalized Difference Vegetation Index (NDVI) across each island and time point. For further details on the metapopulation model, please refer to the Methods section. Data used to produce the maps shown in (a) were made available under Attribution 3.0 Unported (CC BY 3.0)67 and Creative Commons Attribution for Intergovernmental Organisations (CC BY-IGO)68 licenses. The former (CC BY 3.0) licenced the data for Mayotte69, and the latter (CC BY-IGO) licenced the data for the Union of Comoros70, Tanzania71, Madagascar72 and Mozambique73. All presented data was unaltered.
