2012 Dallas County Mosquito Database

Abstract

Background: In 2012, the United States experienced the largest outbreak of West Nile virus (WNV), with the majority of cases and deaths occurring in Texas. Texas reported 1,024 cases of WNV fever, 844 cases of WNV neuroinvasive disease, and 89 deaths throughout the state, with the majority of incidents occurring in Dallas, TX and surrounding areas. Previous studies explored relationships between human cases of WNV and demographic and landscape variables; however, the infection of mosquitoes may better reflect spatial variation in transmission intensity than human cases. In this study, we identified associations between features of the landscape and human population and Culex quinquefasciatus infection with WNV during the 2012 WNV epidemic in Dallas County.

Methods: Using generalized linear mixed models, the minimum infection rate of WNV in Cx. quinquefasciatus was modeled using variables describing the environment and social demographics.

Results: During this epidemic, 25,917 mosquitoes were pooled and tested for WNV, of which 22,156 Cx. quinquefasciatus mosquitoes were identified. Out of 1,634 pools containing at least one Cx. quinquefasciatus mosquitoes, 256 pools (16%) tested positive for WNV. Cx. quinquefasciatus pools accounted for 96% of the positive pools in 2012. Major mosquito and WNV activity occurred between May and September, with a peak in the infection rate during the third week of July (47.7 per 1,000). We found increased probabilities for WNV-positive mosquitoes in north and central Dallas County. Based on a best-fit model generalized linear mixed model, the most significant predictors of the presence of WNV in Cx. quinquefasciatus pools were an index of urbanization (composed of greater population density, lower normalized difference vegetation index, higher coverage of urban land types, and more impervious surfaces), lower elevation, and older populations.

Conclusions: The best-fit model identifies key environmental and demographic factors that play a role in the mosquito’s life cycle and the ability to obtain the virus during the 2012 WNV epidemic in Dallas County, TX. These relationships between the landscape and risk of enzootic transmission help to identify spatial regions of the landscape with highest risk of spill-over to human disease.