Short term prediction of Culex quinquefasciatus population carrying West Nile Virus in Central North Georgia, U.S.A based on the climate variability

摘要

Culex quinquefasciatus, the southern house mosquito, is widely distributed in the southeast United States and has increased the West Nile Virus positive mosquito pools. Especially Georgia and Atlanta area, has been a hotspot of WNV incidence in 2012 with 117 WNV human cases, 6 deaths and 125 WNV positive mosquitos reported by Georgia Department of Public Health. To control mosquito populations and to prevent the disease, understanding this vector-environment relationship is essential. Climatic variability is an important driver of inter-annual WNV transmission risk. To describe the variations in abundance over time and to investigate the correlation between climate forces and inter-annual and seasonal variation of WNV transmission risk, cross-correlation maps (CCMs) were developed for four weeks moving average data from the central north part of the State of Georgia (GA). Using this method, the key antecedent environmental conditions, their timings, and durations were identified which can help improve the ability of developing predictive models of vector abundances. The interval lags with the highest correlation value were selected and were fed into the Poisson regression model and also to Artificial Neural Network (ANN) model for the prediction purposes. In addition, Culex vector abundance 1 to 3 weeks prior to the events were added to both models as predictors. Results of the CCMs and the regression model revealed that cold late winter and early spring decreased mosquito population. Also higher than normal moisture in spring favored Culex mosquitoes carrying WNV in summer. The ANN model predicted the seasonal cycle of mosquito abundance fairly accurately. This model can be used for planning of mosquito control strategies and simulation of mosquito borne diseases by considering climate data as the main driver of WNV risk.