Persistent West Nile virus transmission and the apparent displacement St. Louis encephalitis virus in southeastern California, 2003-2006

Abstract

West Nile virus (family Flaviviridae, genus Flavivirus, WNV) invaded the Colorado Desert biome of southern California during summer 2003 and seemed to displace previously endemic St. Louis encephalitis virus (family Flaviviridae, genus Flavivirus, SLEV, an antigenically similar Flavivirus in the Japanese encephalitis virus serocomplex). Western equine encephalomyelitis virus (family Togaviridae, genus Alphavirus, WEEV), an antigenically distinct Alphavirus, was detected during 2005 and 2006, indicating that conditions were suitable for encephalitis virus introduction and detection. Cross-protective "avian herd immunity" due to WNV infection possibly may have prevented SLEV reintroduction and/or amplification to detectable levels. During 2003-2006, WNV was consistently active at wetlands and agricultural habitats surrounding the Salton Sea where Culex tarsalis Coquillett served as the primary enzootic maintenance and amplification vector. Based on published laboratory infection studies and the current seroprevalence estimates, house sparrows, house finches, and several Ardeidae may have been important avian amplifying hosts in this region. Transmission efficiency may have been dampened by high infection rates in incompetent avian hosts, including Gamble's quail, mourning doves, common ground doves, and domestic pigeons. Early season WNV amplification and dispersal from North Shore in the southeastern portion of the Coachella Valley resulted in sporadic WNV incursions into the urbanized Upper Valley near Palm Springs, where Culex pipiens quinquefasciatus Say was the primary enzootic and bridge vector. Although relatively few human cases were detected during the 2003-2006 period, all were concentrated in the Upper Valley and were associated with high human population density and WNV infection in peridomestic populations of Cx. p. quinquefasciatus. Intensive early mosquito control during 2006 seemed to interrupt and delay transmission, perhaps setting the stage for the future reintroduction of SLEV.

Fig. 1

Twenty-year history of SLEV activity in (A) sentinel chicken flocks and (B) mosquito pools before the arrival of WNV. Data from California Mosquitoborne Encephalitis Virus Surveillance Program summarized from Hui et al. (1999) and Steinlein et al. (2003).

Fig. 2

Map of Coachella Valley showing the locations of fixed mosquito trapping sites and chicken flocks within each geographic zone. Insert shows locations of sites at Zone A along the southern shore of the Salton Sea in Imperial County.

Fig. 3

Changes in rainfall in mm and maximum and minimum temperature in °C recorded at the Mecca Fire Station, Riverside County, CA (33° 32′ N, 119° 59′W, elevation 55 m), January 2003–December 2006. Horizontal line shows the minimal temperature required for WNV replication in Cx. tarsalis mosquitoes.

Fig. 4

Monthly and yearly changes in the maximum elevation of the surface of the Salton Sea in meters below sea level at North Shore, Riverside County, CA (33° 33′ N, 115° 49′ W).

Fig. 5

Geometric mean abundance in Cx. tarsalis and Cx. p. quinquefasciatus (quink) per trap night collected at replicated dry ice-baited traps within each of the seven zones during 2003−2006.

Fig. 6

Cx. tarsalis and Cx. p. quinquefasciatus (quink) WNV infection rates per 1,000 females tested calculated by the maximum likelihood method of Biggerstaff and the cumulative number of sentinel chicken seroconversions per flock (chick) within each zone plotted as a function of time in months within each year.

Fig. 7

Sentinel chicken seroconversions to WNV per flock per year within each of seven zones.

Fig. 8

(A) Total avian sera positive for SLEV, WNV and unidentified flavivirus (FLAV) plotted as a function of month of collection. (B) Monthly avian seroprevalence rates against WNV among year round residents of Coachella Valley.