Epidemiology of West Nile Virus: A Silent Epiornitic in Northern Delaware in 2007 without Associated Human Cases

We performed a 2-year longitudinal study (2006–2007) of West Nile virus (WNV) infections in wild birds, mosquitoes, and sentinel chickens at 6 WNV-endemic sites in northern Delaware. We determined virus infection rates of Culex pipiens and other mosquito vectors as well as seroprevalence and antibody titers of amplifying hosts. Endemicity status varied widely among the 6 sites based on 3 criteria—mosquito infections, sentinel chicken seropositivity, and wild bird seropositivity. A highly endemic site would display at least 2 of the 3 criteria during each year of the study, while a site with just 1 positive criterion was considered to have low endemicity. Culex pipiens was the principal vector detected at 2 highly endemic sites in 2006 vs. 1 site in 2007. However, in 2006, we also found 2 other WNV-positive vector species as well as an unidentifiable Culex species at 1 highly endemic site, suggesting increased activity at the end of the 1st year of the study. Wild birds were early indicators of WNV at highly endemic sites in mid-July to early August of both 2006–2007. Mosquitoes were positive in mid- to late August, appearing concurrently with seroconverted sentinel chickens, with wild resident birds appearing ∼4 wk prior to those indicators. Of birds tested with n ≥ 9, Northern cardinals had the highest seropositivity rates (47%) followed by Carolina wrens (19%), house sparrows (13%), American robins (13%), tufted titmice (11%), and gray catbirds (9%). The overall seropositive rates in trapped birds increased from 5.0% in 2006 to 20.0% in 2007, while the geometric mean titers of all positive birds increased from 1∶34 to 1∶47 during the comparable periods. Based on these results, we suggest that an epiornitic in birds occurred in 2007, but that greatly reduced abundance of mosquito vectors caused by an extreme drought largely precluded human infection.