Predicting West Nile Virus Mosquito Infection in Long Island, New York through spatial modeling

What is it about?

The research confirmed that patterns of warm weather and low spring and summer rainfall are associated with increased West Nile incidence in mosquitoes. Woody and emergent wetlands provided the ecosystem service of vector control, by providing habitat for mosquito predators that eat larvae and reduce the overall adult mosquito population in the area. The number of nearby septic systems had a positive relationship with West Nile incidence, providing further evidence that onsite sewage treatment systems can, in some circumstances, encourage mosquito breeding and the spread of disease. The statistical approach was able to quantify both spatial and temporal variability as well as the impact of the predictor variables, leading to a robust model that correctly predicted West Nile Virus mosquito incidence approximately eighty percent of the time.

Featured Image

Why is it important?

West Nile Virus has become an endemic disease in New York City and nearby Long Island in the last two decades. It is transmitted by the bite of an infected mosquito and can cause symptoms from fever and fatigue to brain damage or death. Because treatment is difficult, the best way to prevent infections is by targeting mosquito control efforts to places the disease is likely to spread. We used a combination of environmental variables such as land use type, type of plant cover, and rainfall along with spatial and temporal Bayesian modeling to accurately forecast areas of high West Nile Virus infection in Suffolk County, New York mosquitoes. The number of nearby septic systems was a variable of particular interest, because Suffolk County has an unusual density of unsewered residences and on-site sewage treatment systems are suspected to contribute to mosquito breeding habitat.