West Nile Virus Illinois

NON-TECHNICAL SUMMARY: The hallmarks of West Nile Virus (WNV) in the U.S. have been: 1) Rapid range expansion, 2) High mortality in crows and high prevalence in many other avian species, 3) Over-wintering and re-emergence early in the season at multiple sites, and 4) Transmission cycle similar to St. Louis encephalitis virus with the virus amplified in avian hosts. Our investigations will address two areas of applied research: 1) The population dynamics and feeding behavior of potential mosquito vectors of WNV, and 2) The seasonal interactions of vector mosquitoes and avian hosts.

<P>
APPROACH: We will compare seasonal dynamics of key mosquito species by collecting mosquitoes at selected sites using 5 methods. Multiple sampling methods will be used because the efficacy of each method varies by mosquito species, habitat, and physiological state. At each site, mosquito collections will be conducted for 1-3 days before and during the bird collections (Objective 2) in the same area. Collection data will be analyzed, in conjunction with meteorological and ecological variables, to determine seasonal trends in regional abundance, indices of mosquito species association, and correlation of key events and temporal events to environmental variables. We will determine the relative abundance of key bird species from northern to southern Illinois during May to September each year, using surveys in conjunction with bird and mosquito trapping. We will analyze temporal overlap in host and mosquito abundance and development. Results from the mosquito and host sampling and serology will be entered into a real-time, statewide database, which will allow predicting disease outbreaks based on samples of hosts and mosquitoes, as well as human cases of WNV. This software will allow managing spatial data, and allow user-generated maps and tables of all data via the internet. Sampling and models developed will help estimate the magnitude of the disease, which will be critical for mitigating the incidence of vectors positive for WNV, using appropriate IPM methods in both urban and rural areas, thus helping avert impacts of infection in humans and animals.
<P>
PROGRESS: 2004/06 TO 2007/06<BR>
In cooperation with several mosquito abatement districts in Cook Co., IL, approximately 11,000 mosquito pools (nearly 40,000 mosquitoes) were tested for West Nile virus (WNV) in the past year. Nearly 1,800 pools, primarily Culex restuans and Cx. pipiens, were positive. Approximately 400 birds were tested for WNV-antibodies using ELISA, and more than 17% were positive. Infection rates in Culex species collected in Cook County were relatively high over extended periods in 3 of the past 5 years. Prevalence ranged from 4-14% WNV positive for 2-4 weeks of the season. Temperature appeared to be the major regulator of transmission intensity and the increase in infected mosquito pools paralleled the increase in Cx. pipiens abundance. Studies comparing temporal change in vector abundance, infection rate, and human cases by onset date, showed that density of infected mosquitoes was the best indicator of risk to humans, and changes in density tended to precede and parallel changes in human cases by 2-3 weeks. Spring bird counts and breeding bird surveys throughout Illinois showed declines in populations of corvids and some cavity nesters in areas with high WNV transmission in mosquitoes and humans. Birds contaminated feeders with low levels of WNV. In Cook Co., abundance of Cx. pipiens was greater than in Central Illinois and numbers tended to build up faster in Cook Co., which may reflect higher temperatures in suburban "heat islands" and more underground winter refuges. Recent tests of samples suggest an increase in the discontinuity between VecTests and TaqMan, possibly reflecting local variation in WNV antigenic properties and nucleic acid sequences. Bloodmeal analysis and seroprevalence studies implicated the same avian species as major hosts; American robins, northern cardinals, mourning doves, and house sparrows. The rank in seroprevalence, however, did not necessarily agree with the rank in blood meals. American robins, for example, outnumbered the other species in the blood meal analysis, but not in seroprevalence. The proportion of mammal vs. bird bloods varied spatially based on abundance of available hosts in the area. Mosquito traps near or in horse stables, cow barns, or pig stables had almost 20% feeding on mammals, whereas blooded mosquitoes caught in residential/campus areas had less than 7% mammal blood. One crow blood and one human blood were found in 50 blood meals identified to species, indicating these species were low preference. In Central Illinois, both molecular analysis and morphology of male reproductive structures showed that hybrids of Cx. pipiens with Cx. quinquefasciatus occur in Champaign-Urbana in August and September. This area is usually considered outside the northern boundary of the hybrid zone. The start of mosquito breeding was early, March and April 2007 in Central Illinois, and early season Cx. restuans were RT-PCR positive. This suggests overwintering Cx. restuans may be the initial source of enzootic transmission. WNV was not detected in overwintering Cx. pipiens (0 positives out of approx. 3,500 mosquitoes collected from stormwater tunnels).
<P>
IMPACT: 2004/06 TO 2007/06<BR>
These data challenge several previous conclusions: 1) No late season increase in mammal feeding was found as reported for the East Coast. 2) The blood-feeding pattern must be expressed in terms of both temporal change and spatial availability of hosts; proportion of mammal feedings varies. 3) Decline in WNV transmission tends to begin in early Aug., corresponding to the end of breeding season for common urban mosquito species. 4) Crows are not essential for WNV amplification. 5) Extended collection of egg rafts in Central Illinois when photoperiod is reduced to > 14 hours of light is probably due to the presence of Cx. pipiens complex hybrids that lack an ovarian diapause. A similar late season oviposition pattern was noted for Cook Co, but is unlikely due to Cx. quinquefasciatus hybrids because the location is beyond the hybrid zone. 6) These studies suggest that density of infected mosquitoes is a better indicator of risk than the number of infected biting mosquitoes x mammal feeding preference x vector competency. 7) These studies provide first evidence that Cx. restuans overwinters the virus in Illinois. 8) Preliminary data suggest the vast underground stormwater/sewage system is responsible for the timing and abundance of Cx. pipiens in Cook Co. These studies established basics of the urban transmission cycle in central and northern IL, and testable hypotheses can now be developed to elucidate critical determinants of transmission and when intervention might be most efficaceous.