Landscape ecology of arboviruses in southern California: temporal and spatial patterns of vector and virus activity in Coachella Valley, 1990-1992.

Consistent temporal and spatial patterns in the activity of Culex tarsalis Coquillett and western equine encephalomyelitis (WEE) and St. Louis encephalitis (SLE) viruses were delineated that were useful in developing a stratified surveillance program. Vernal increases in Cx. tarsalis abundance typically were associated with flooding of saline marshes along the north shore of the Salton Sea and were followed 6-8 wk later by the onset of WEE and SLE virus activity. Viruses then spread to managed marsh (duck club) and agricultural habitats in the Whitewater Channel flood plain and, depending upon the intensity of amplification, to agricultural and residential areas in the more elevated northwestern portion of the valley. Mean annual Cx. tarsalis abundance was correlated inversely with elevation and distance from the Salton Sea. Abundance was greatest at managed marsh habitats. Although spatially correlated with vector abundance among sites, virus transmission rates to sentinel chickens were asynchronous temporally with vector abundance. Seroconversion rates were related to flock location but not flock size (10 versus 20 chickens). Human cases were not detected during the study period, despite elevated transmission rates of both WEE and SLE viruses to sentinel chickens positioned in peridomestic habitats.

Potential effect of global warming on mosquito-borne arboviruses.

If global warming occurs in California, daily mean temperatures may increase by 3 to 5 degrees C, precipitation patterns will change, and sea level may rise 1 m. Studies were done on effect of temperature changes on survival of Culex tarsalis Coquillett, the primary vector of western equine encephalomyelitis (WEE) and St. Louis encephalitis (SLE) viruses, in two regions where temperatures differed by 5 degrees C. Daily mortality of adult vectors increased by 1% for each 1 degree C increase in temperature. At 25 degrees C, only 5% of Cx. tarsalis survived for 8 or more days, the time required for extrinsic incubation of these viruses. Extrinsic incubation times for these viruses shortened when temperatures were increased from 18 to 25 degrees C. WEE virus infection was modulated and transmission decreased at 32 degrees C. If temperatures in the warmer region increase by 5 degrees C, WEE virus may disappear and SLE virus would persist. In the cooler region, a 5 degrees C increase would decrease vector survivorship and virus activity in midsummer. In North America, epidemics of WEE have prevailed above a 21 degrees C isotherm and those of SLE below this isotherm. With global warming, epidemics of these viruses could extend into currently unreceptive northern areas. WEE virus would disappear from more southern regions. Geographic distribution of vector, human, and animal populations could be altered. North America could become more receptive to invasion by tropical vectors and diseases.

Ecology of mosquitoes and St. Louis encephalitis virus in the Los Angeles Basin of California, 1987-1990.

St. Louis encephalitis (SLE) virus has become established in the Los Angeles Basin of California, where it most likely is maintained by horizontal transmission among Culex tarsalis Coquillet and passerine birds in park-riparian habitat. Viral transmission also was detected at low levels in residential habitat by the infrequent seroconversion of sentinel chickens. SLE virus activity was documented in all months except March and April. Cx. tarsalis was incriminated as the primary vector, based on elevated field infection rates (0.17 infected females per 1,000 tested), population abundance concomitant in time and space with increased SLE virus activity, vector competence, and vectorial capacity. Culex stigmatosoma Dyar also may be important in the horizontal maintenance of SLE virus because this species was an efficient laboratory vector, fed frequently on passeriform birds, and had field infection rates (0.15 females per 1,000) comparable with Cx. tarsalis. However, difficulty in trapping adult females precluded the assessment of Cx. stigmatosoma because abundance may have been underestimated and comparatively few females were tested for virus. Although Cx. quinquefasciatus Say was the most abundant species in residential habitats (greater than 85% of total catch at CO2 traps) where it occasionally fed on humans (0.6% of total tested), this species was considered to be of secondary importance because field infection rates were low (0.03 females per 1,000) and few females became infected and transmitted SLE virus in laboratory vector competence experiments. Although Culex erythrothorax Dyar fed on humans in marsh habitat (5% of total tested), this species infrequently fed on birds (12%) and was refractory to SLE virus infection in vector competence experiments; therefore, it was not considered a vector of SLE virus. Future SLE virus surveillance and mosquito control programs should be directed toward Cx. tarsalis populations.

Mosquito and arbovirus ecology in southeastern California, 1986-1990.

Mosquito abundance and western equine encephalomyelitis (WEE) and St. Louis encephalitis (SLE) virus activity were monitored in five valleys in southeastern California from June 1986 through April 1990 to study virus overwintering and possible dissemination from south to north along geographically defined corridors. Culex tarsalis Coquillett predominated in CO2 trap collections and was the only species repeatedly infected with WEE and SLE viruses. Abundance peaked during April-May and August-October. WEE virus infections in Cx. tarsalis generally were detected after the spring peak and were followed approximately 1 mo later by seroconversions in sentinel chickens. SLE virus infections occurred later in the summer but before the fall peak in Cx. tarsalis abundance. Peak Cx. tarsalis abundance occurred when monthly temperatures averaged 25 degrees C, whereas virus infections generally were detected most frequently when temperatures exceeded 29 degrees C. Although the spring increase in Cx. tarsalis abundance occurred earlier in southern valleys, the onset of virus activity was variable among valleys and did not follow a south to north progression.

Ecological observations on the 1989 outbreak of St. Louis encephalitis virus in the southern San Joaquin Valley of California.

Temporal and spatial patterns of St. Louis encephalitis (SLE) virus transmission were compared at permanent study areas in the southern San Joaquin Valley during years with low (1988 and 1990) and elevated (1989) viral activity. During 1989 and 1990, virus appeared first at sentinel chicken flocks exhibiting low to moderate seroconversion rates at the end of the previous season. This finding, and the early season seroconversion of sentinel chickens at a marsh habitat on 5 March and 2 April 1990, circumstantially indicated that SLE virus may have overwintered on the valley during the winters of 1988-1989 and 1989-1990. The mechanism of overwintering was not elucidated further, because virus could not be isolated from overwintering adult mosquitoes or from immatures collected during the spring. An outbreak of 26 confirmed SLE cases occurred in 1989 during a drought year (rainfall 50% of normal) and followed a spring with elevated temperatures (1.7-3.4 degrees C above normal) and Culex tarsalis Coquillett abundance. Cx. tarsalis was the primary vector, being most abundant during the virus amplification period in early summer and most frequently infected (70 SLE virus positive pools/329 tested). SLE virus also was detected in Culex quinquefasciatus Say (14/65) and Cx. stigmatosoma Dyar (1/4); however, both species were distributed focally and increased in abundance only after widespread seroconversions had occurred in sentinel chickens. Increased virus activity during 1989 was not accompanied by marked changes in vector susceptibility or in SLE virus infectivity for mosquitoes. Decreased virus activity in the Bakersfield area during 1990 could not be attributed to immunity in passeriform birds, because a small seroprevalence survey indicated that few adult birds had antibodies to SLE virus.

Population dynamics of adult Culex mosquitoes (Diptera: Culicidae) along the Kern River, Kern County, California, in 1990.

The temporal and spatial abundance, dispersal, survivorship, and density of Culex tarsalis Coquillett and Cx. quinquefasciatus Say populations were studied in riparian, agricultural, and residential habitats along the Kern River bed during the drought year of 1990. The temporal abundance of both species was related to cotton agricultural practices and peaked during intensive irrigation after cultivation was terminated in July. Cx. tarsalis peaked in abundance 2 wk earlier than Cx. quinquefasciatus, perhaps because of the advantage of autogenous oviposition, which shortened generation time. Although host-seeking females of both species were most abundant in the riparian habitat, more Cx. quinquefasciatus than Cx. tarsalis were collected in the residential habitat. Marked females released within the riparian habitat were recaptured most frequently within 1 km of the release point; however, flights as far as 12.6 km were documented within the 180-km2 study area. On average, marked Cx. quinquefasciatus dispersed farther and more rapidly than Cx. tarsalis and were recaptured more frequently within the residential habitat. Survivorship, estimated horizontally from the female recapture rate, ranged from 0.60 in May to 0.79 in July for Cx. tarsalis and from 0.74 in September to 0.84 in July for Cx. quinquefasciatus. The estimation of survivorship vertically from the parity rate was complicated by elevated autogeny rates in Cx. tarsalis and by the low parity rate in Cx. quinquefasciatus. Cx. tarsalis population density ranged from 125 females per km2 in May to 65,500 per km2 in August, and was well correlated with relative abundance. Collectively, these ecological data indicated that Cx. tarsalis may be important in disseminating arboviruses within the riparian habitat, but that Cx. quinquefasciatus may be important secondarily by disseminating virus from the riparian habitat to the adjacent residential habitat.

Influence of vegetation on carbon dioxide trap effectiveness for sampling mosquitoes in the Sierra Nevada foothills of Kern County, California.

The effect of vegetation on sampling Culex tarsalis, Cx. quinquefasciatus and Aedes nigromaculis by CO2 traps was evaluated at an intermittent stream habitat at the base of the Sierra Nevada foothills. Carbon dioxide traps were spaced along a 450 m transect perpendicular to Poso Creek to determine female attraction to traps placed in 5 different vegetation substrates: 1) open hilltop with sparse growth of grasses and saltbush, 2) open pasture with sparse growth of saltbush, 3) peripheral understory of mule fat, 4) shaded understory of mule fat, and 5) open canopy 5 m above ground in willow and cottonwood trees. Most host-seeking Cx. tarsalis and Cx. quinquefasciatus females were collected within the open canopy and peripheral understory. Host-seeking Ae. nigromaculis females were collected predominately in the open pasture and within the peripheral understory. The association between CO2 trap catch size and vegetation suggested a relationship between the host-feeding patterns and associated hunting strategies of these bird and mammal feeding species.

Mark-release-recapture studies with Culex mosquitoes (Diptera: Culicidae) in Southern California.

Cohorts of Culex adults were marked uniquely with date- and site-specific fluorescent dust colors and were released at centrally located residences and at peripheral breeding sources to study population dispersal, size, additions, and deletions. The recapture rate of Cx. quinquefasciatus females was higher and the mean distance dispersed was lower in residential than in agricultural or park habitats. Dispersal was associated with host-seeking activity and ranged from 0.6 to 1.0 km/d. Survivorship ranged from 0.65 to 0.84 per day, and population density ranged from 36,612 to 671,634 females per km2. The sampling efficiency of CO2-baited traps in residential habitats increased coincidentally with increasing population density. Gravid traps were most effective in residential habitats where there were few competitive oviposition sites. Teneral Cx. stigmatosoma were extremely dispersive, and few marked females were recaptured. Unmarked females were more abundant at CO2-baited traps in residential habitats than at traps near productive peripheral breeding sources. Few Cx. tarsalis were released, and the recapture rate in residential habitats was low when compared with rural sites.

Persistence of mosquito-borne viruses in Kern County, California, 1983-1988.

The persistence of arboviruses was studied from 1983 to 1988 in mixed agriculture, marsh, riparian, and foothill habitats in Kern County, CA. Western equine encephalomyelitis (WEE) virus was isolated frequently during 1983 from Culex tarsalis and Aedes melanimon and was detected by the seroconversion of sentinel chickens. WEE virus then disappeared, even though vector competence studies during 1984-1986 showed that Cx. tarsalis was able to transmit WEE virus. St. Louis encephalitis (SLE) virus was detected sporadically in 3 of the 6 years of the study by isolation from Cx. tarsalis and/or by sentinel chicken seroconversion. When mosquito pools were screened for virus in suckling mice, Turlock (TUR) and Hart Park (HP) viruses were isolated from Cx. tarsalis during each summer. Vertical transmission of HP was indicated by the isolation of virus from a pool of male Cx. tarsalis. California encephalitis (CE) virus was isolated repeatedly from host-seeking Ae. melanimon females, males, and adults reared from field-collected immatures, verifying vertical transmission in nature. Horizontal transmission of CE virus among both jackrabbits (Lepus californicus) and desert cottontails (Sylvilagus auduboni) appeared to amplify Ae. melanimon infection rates during the summer of 1985, but elevated herd immunity depressed infection rates during 1986. Thus, CE, HP, and TUR viruses persisted in Kern County, while WEE virus appeared to become extinct and required reintroduction. The sporadic occurrence of SLE virus activity remains unexplained, but its persistence may require both vertical transmission and reintroduction.

Mosquito bionomics and the lack of arbovirus activity in the Chino area of San Bernardino County, California.

Mosquito bionomics, including temporal abundance, metabolic status, and blood-feeding patterns, and arbovirus activity were studied at representative dairy and residential habitats in Chino, Calif., during 1987 and 1988. Host-seeking Culex tarsalis Coquillett and Cx. quinquefasciatus Say females were more abundant and appeared earlier each season at traps near dairy breeding sources than at traps in residential areas, whereas Cx. stigmatosoma Dyar, which also breeds in dairy effluent, was most abundant at residential traps. All three Culex species remained gonotrophically active throughout winter and did not appear to enter diapause. Dairy cows diverted the normally ornithophagic Cx. tarsalis and Cx. quinquefasciatus females from avian hosts in dairy, but not in residential, habitats. Cx. stigmatosoma fed almost exclusively on birds. The diversion of host-seeking Cx. tarsalis females to bovine hosts and reduced abundance because of mosquito abatement may have combined to reduce the receptivity of the Chino area to St. Louis encephalitis virus.

Temporal variations in the susceptibility of a semi-isolated population of Culex tarsalis to peroral infection with western equine encephalomyelitis and St. Louis encephalitis viruses.

A semi-isolated population of Culex tarsalis in Kern County, CA was found to vary significantly in its seasonal and yearly susceptibility to peroral infection with western equine encephalomyelitis (WEE) and St. Louis encephalitis (SLE) viruses during the breeding seasons of 1975-1981. Female cohorts of the population were significantly more resistant to WEE virus from 1975 through 1977 than from 1978 through 1981. On the average, females were 40 times more susceptible to WEE virus during May than during August. Increases in resistance correlated significantly with the number of days accrued from April through June, when daily ambient air temperatures equaled or exceeded 26.7 degrees C or 32.2 degrees C, respectively. Analyses of data on WEE viral susceptibility, temperature, and rainfall for the 7 year period suggested that Cx. tarsalis should be most susceptible to infection in years with wet winter-early spring periods followed by cool springs. However, inconsistencies in the expression of WEE viral susceptibility during mid- to late summer of some years indicated that other unidentified extrinsic factors, in addition to temperature and rainfall, may induce changes during preimaginal development that affect the peroral susceptibility of adult females. Seasonal and yearly changes in the susceptibility of the Cx. tarsalis population to per os infection with SLE virus did not necessarily occur at the same time as those observed with WEE virus and did not appear to correlate with changes in ambient air temperature. Thus, extrinsic factors that affect the expression of viral susceptibility of Cx. tarsalis, and perhaps the genetic basis of viral susceptibility, apparently differ for WEE and SLE viruses.

Studies on the seasonality of Culiseta inornata in Kern County, California.

The seasonal abundance of adult Culiseta inornata was markedly bimodal in the San Joaquin Valley of Kern County, California, with minima observed during both summer and midwinter. Larvae were abundant in most surface water habitats during winter, but could not be found during summer. The occasional collection of females during summer indicated the persistence of an adult population. The midwinter decrease in adult abundance was attributed to the progressive mortality of the autumnal cohort and delayed emergence due to cold water temperature. Reproductive diapause was not induced experimentally when field or laboratory populations were exposed as larvae, pupae or adults to simulated summer or winter photoperiod and temperature regimens. In comparison, Culex tarsalis readily entered a winter diapause when concurrently exposed to simulated winter conditions. The aestivation, and perhaps hibernation, of reproductively quiescent females makes Cs. inornata theoretically attractive as a maintenance host of encephalitis viruses, while the bimodal seasonality of host-seeking activity defines periods when Jamestown Canyon virus may be transmitted horizontally.

Studies on autogeny in Culex tarsalis: 3. Life table attributes of autogenous and anautogenous strains under laboratory conditions.

The reproductive biology and life table attributes of autogenous and anautogenous strains of Cx. tarsalis which were selected from the same parent colony were compared under laboratory conditions. Autogenous mosquitoes required 1 day longer to complete immature development, but oviposited 1 to 2 days earlier than anautogenous mosquitoes. Autogenous females readily imbibed blood meals from restrained chickens if ovarian maturation had not progressed to Christophers' Stage III. Wing length and life expectancy were not significantly different between strains; however, autogenous females laid a significantly smaller number of eggs per raft during initial oviposition than anautogenous females. Egg raft size did not differ significantly between strains during subsequent ovipositions resulting in similar net reproductive rates (Ro). Earlier oviposition and a comparable Ro resulted in a greater intrinsic rate of increase (rm) and birth rate (b) for autogenous than anautogenous cohorts. Thus, highly autogenous populations would be able to exploit newly created surface water breeding sources more rapidly than highly anautogenous populations. However, highly autogenous populations probably would not be able to transmit a horizontally maintained arbovirus as efficiently as anautogenous populations, since autogenous females imbibe their initial blood meal later in life than anautogenous females.

Air conditioning and television as protective factors in arboviral encephalitis risk.

In California, the advent of television and air conditioned housing has coincided with a general decrease in mosquito-borne viral encephalitis cases in humans in the past 25 years. During this same period, levels of vector populations, virus activity in vectors, birds, and horses has been high at times. Air conditioning and television encourage persons to remain indoors during summer evenings, the primary time when infected Culex tarsalis transmit western equine encephalomyelitis and St. Louis encephalitis viruses. The attack rates for these 2 diseases in 33 counties in the central valley of California, 1945-1982, were compared with the prevalence of household air conditioning and television. Encephalitis rates were negatively associated with television ownership in both cross-sectional and longitudinal analysis. Moreover, individual counties which experienced the most rapid increase in household television ownership had the greatest decline in encephalitis rates. A telephone survey conducted in Kern County revealed that air conditioner and television utilization times corresponded closely to the feeding times of Cx. tarsalis, and respondents indicated a preference to remain indoors during this time because of these appliances. It is concluded that changed behavioral patterns may protect from vector-borne diseases and be complementary to vector control programs.

The influence of constant versus fluctuating water temperatures on the preimaginal development of Culex tarsalis.

Culex tarsalis larvae derived from field-collected egg rafts were reared outdoors in a deep pond with constant water temperature or a shallow pond with fluctuating temperature. No significant differences were found between ponds in the proportion of larvae surviving to adults, female autogeny rates or wing lengths of adults. The average time from 1st instar to adult emergence ranged from 8 days at 31 degrees C to 16 days at 17 degrees C. Males developed faster than females, and both sexes always emerged faster from the shallow pond than from the deep one. However, the temperature-adjusted mean emergence times were equal, and normalized cumulative development curves, fitted to a Weibull function, were identical for the 2 ponds.

Population dynamics of some Pakistan mosquitoes: changes in adult relative abundance over time and space.

Time series and spatial changes in the relative abundance of 14 mosquito species were described from weekly or biweekly collections at nine localities in Punjab province, Pakistan, from January 1976 to June 1980. Comparisons between indoor aspirator catches and outdoor mechanical sweeper collections enabled mosquito resting habits to be classified as: (1) endophilic (Anopheles culicifacies, An. fluviatilis, An. stephensi, An. subpictus); (2) partially exophilic (An. annularis, An. pulcherrimus, An. nigerrimus, Culex bitaeniorhynchus, Cx, pseudovishnui, Cx, quinquefasciatus, Cx. tritaeniorhynchus), or exophilic (Cx. fuscocephala, Aedes caspius, Mansonia uniformis). Temporal abundance patterns were grouped by seasonality, overwintering strategies and the magnitude of fluctuation. Seasonal patterns were: (1) unimodal-spring (Cx. quinquefasciatus, Ae. capius); (2) unimodal-monsoon (An. nigerrimus, An. subpictus, Cx. bitaeniorhynchus, Cx. fuscocephala); (3) bimodal-spring dominant (An. annularis, An. culcifacies, An. stephensi), and (4) bimodal-monsoon/post-monsoon dominant (An. fluviatilis, An. pulcherrimus, Cx. pseudovishnui, Cx. tritaeniorhynchus, Ma. uniformis). Mosquito overwintering strategies included: (1) adults with slowed reproductive activity (An. annularis, An. culicifacies, An. fluviatilis, An. pulcherrimus, An. stephensi, Cx. quinquefasciatus); (2) females with interrupted reproductive activity (An. nigerrimus, Cx. fuscocephala, Cx. pseudovishnui, Cx. tritaeniorhynchus); (3) immature stages (Ae. caspius, Ma. uniformis) and (4) annual extinction and re-introduction (An. subpictus). The magnitude of seasonal change was classified by the number of standard deviations from the overall mean exhibited by the annual maxima or minima: (1) stationary, less than 1 standard deviation (An. culicifacies, An. fluviatilis, Cx. bitaeniorhynchus, Ae. caspius), (2) fluctuating moderately, one to two standard deviations (An. annularis, An. nigerrimus, An. pulcherrimus, An. stephensi, Cx. fuscocephala, Cx. quinquefasciatus, Ma. uniformis) and (3) markedly fluctuating greater than two standard deviations (An. subpictus, Cx. pseudovishnui, Cx. tritaeniorhynchus). The seasonal increases of An. culicifacies and An. stephensi abundance did not closely parallel the periods of maximal malaria transmission. Temporal patterns exhibited by several culicine species indicated that the period of greatest West Nile virus activity would probably occur during the post-monsoon season.

Aerial adulticiding for the suppression of Culex tarsalis in Kern County, California, using low volume propoxur: 2. Impact on natural populations in foothill and valley habitats.

A low volume formulation of propoxur wettable powder suspended in larvicidal oil was evaluated on 4 occasions in Kern Co., California during 1983 using fixed and rotary wing aerial application systems. Culex tarsalis abundance was suppressed significantly by all sprays, although reduction below the virus maintenance threshold of 30 females per trap night was achieved only at one foothill site. Western equine encephalomyelitis virus minimum infection rates decreased significantly after serial applications on 2 occasions at a valley site. However, virus persisted in the spray zone at minimum infection rates of greater than 1 per 1000 Cx. tarsalis females tested and transmission of virus to sentinel chickens continued. The parity rates were reduced significantly at 2 semi-isolated foothill sites, but not at a valley site where elevated autogeny rates increased the reproductive age of the host-seeking population. Spraying during late afternoon by helicopter resulted in better control than early morning applications by fixed wing aircraft at a valley site.