Statistical modeling of the abundance of vectors of West African Rift Valley fever in Barkédji, Senegal.

Rift Valley fever is an emerging mosquito-borne disease that represents a threat to human and animal health. The exophilic and exophagic behavior of the two main vector in West Africa (Aedes vexans and Culex poicilipes), adverse events post-vaccination, and lack of treatment, render ineffective the disease control. Therefore it is essential to develop an information system that facilitates decision-making and the implementation of adaptation strategies. In East Africa, RVF outbreaks are linked with abnormally high rainfall, and can be predicted up to 5 months in advance by modeling approaches using climatic and environmental parameters. However, the application of these models in West Africa remains unsatisfactory due to a lack of data for animal and human cases and differences in the dynamics of the disease emergence and the vector species involved in transmission. Models have been proposed for West Africa but they were restricted to rainfall impact analysis without a spatial dimension. In this study, we developed a mixed Bayesian statistical model to evaluate the effects of climatic and ecological determinants on the spatiotemporal dynamics of the two main vectors. Adult mosquito abundance data were generated from July to December every fortnight in 2005-2006 at 79 sites, including temporary ponds, bare soils, shrubby savannah, wooded savannah, steppes, and villages in the Barkédji area. The results demonstrate the importance of environmental factors and weather conditions for predicting mosquito abundance. The rainfall and minimum temperature were positively correlated with the abundance of Cx. poicilipes, whereas the maximum temperature had negative effects. The rainfall was negatively correlated with the abundance of Ae. vexans. After combining land cover classes, weather conditions, and vector abundance, our model was used to predict the areas and periods with the highest risks of vector pressure. This information could support decision-making to improve RVF surveillance activities and to implement better intervention strategies.