Mosquito catch in New Jersy light traps (NJLTs) has been declining in recent years, compromising the sensitivity of the California mosquito monitoring program. Centers for Disease Control traps (CDCTs) operated without light and augmented with dry ice have been considered for replacement or augmentation. To provide information on comparative sensitivity and ability to measure abundance over time and space, catch of mosquitoes in NJLTs was compared to catch in CDCTs operated concurrently at 8-10 sites within the Coachella Valley, Kern, San Joaquin County, and Sacramento-Yolo Mosquito and Vector Control Districts. The CDCTs always collected more female mosquitoes than did NJLTs; however, differences in sensitivity varied markedly over time and space precluding the calculation of a universal conversion factor. Regressions of the catch of female Culex tarsalis in CDCTs as a function of catch in NJLTs within districts indicated that the slopes varied markedly, again precluding the derivation of a universal function. Therefore, we recommend that mosquito surveillance programs replace or supplement NJLTs with systematically operated CDCTs to enhance sampling sensitivity for females of most mosquito species. However, both trap types should be operated concurrently at several sites within each district to derive regression functions to convert historical relative abundance data from NJLTs to equivalent counts in CDCTs for retrospective analyses.

Temporal and spatial changes in the enzootic activity of western equine encephalomyelitis (WEE) and St. Louis encephalitis (SLE) viruses were monitored at representative wetland study sites in the Coachella, San Joaquin, and Sacramento valleys of California from 1996 to 1998 using three methods: (1) virus isolation from pools of 50 host-seeking Culex tarsalis Coquillett females, (2) seroconversions in flocks of 10 sentinel chickens, and (3) seroprevalence in wild birds collected by mist nets and, grain baited traps. Overall, 74 WEE and one SLE isolates were obtained from 222,455 Cx. tarsalis females tested in 4,988 pools. In addition, 133 and 40 seroconversions were detected in 28 chicken flocks, and 143 and 27 of 20,192 sera tested from 149 species of wild birds were positive for antibodies to WEE and SLE, respectively. WEE was active in all three valleys, whereas SLE only was detected in Coachella Valley. Seroconversions in sentinel chickens provided the most sensitive indication of enzootic activity and were correlated with seroprevalence rates in wild birds. Avian seroprevalence rates did not provide an early warning of pending enzootic activity. in chickens, because positive sera from after hatching year birds collected during spring most probably were the result of infections acquired during the previous season. Few seroconversions were detected among banded recaptured birds collected during spring and early summer. Age and resident status, but not sex, were significant risk factors for wild bird infection, with the highest seroprevalence rates among after hatching year individuals of permanent resident species. Migrants (with the exception of mourning doves) and winter resident species rarely were positive. House finches, house sparrows, Gambel's quail, California quail, common ground doves, and mourning doves were most frequently positive for antibodies. The initial detection of enzootic activity each summer coincided closely with the appearance of hatching year birds of these species in our study areas, perhaps indicating their role in virus amplification. Bird species most frequently positive roosted or nested in elevated upland vegetation, sites where Cx. tarsalis host-seeking females hunt most frequently. These serosurveys provided important background information for planned host competence and chronic infection studies.

The effectiveness of New Jersey (NJ) light, dry ice baited, and gravid female traps for collecting adult mosquitoes was compared at representative habitats in the Coachella, San Joaquin, and Sacramento valleys and the Los Angeles basin of California. The NJ light traps effectively sampled Anopheles freeborni, Culex tarsalis, Psorophora columbiae, and several Aedes when abundance was high in rural areas with minimal competitive illumination. Dry ice-baited encephalitis virus surveillance or CDC style traps collected significantly more females of most species at most localities than did NJ light traps, regardless of background illumination. The Cummings modification of the Reiter gravid female trap baited with a bulrush (Schoenoplectus) infusion was the best method for collecting Culex pipiens complex females in most habitats. In the Los Angeles basin, gravid traps baited with bulrush infusion collected, on average, 4.5 times more Culex quinquefasciatus females than did traps baited with the Reiter infusion. The bulrush infusion in combination with the Cummings trap design seemed to provide resting site cues and collected males as well as empty and bloodfed females. Mosquito surveillance programs in California should include the systematic operation of dry ice-baited and gravid female traps to improve surveillance sensitivity for selected species in appropriate habitats.

The temporal and spatial distribution of western equine encephalomyelitis virus (WEEV) in California is described during 1993 and 1994 based on seroconversions in sentinel chickens and infected pools of Culex tarsalis Cog. A single horse and no human cases were reported despite widespread enzootic transmission in the Sacramento Valley. During 1993, elevated sentinel chicken seroconversion rates in the Sacramento Valley were related to cool vernal temperature, above normal rainfall and river discharge, and elevated infection rates in Cx. tarsalis. Western equine encephalitis virus appeared to overwinter in the Sacramento Valley and was detected again in 1994; however, warm vernal temperatures and reduced water availability limited the intensity and distribution of transmission to sentinel chickens. The absence of human cases and possible mechanisms for WEEV persistence are discussed.