Weather & climate forecasts

Weather & climate forecasts

FIGURE 3-3 TOPS system brings ground and remote measures of climate into ecological models to monitor and forecast risk. SOURCE: NASA (2007).

abundance; (Reisen et al., 2008a) and outbreaks of arboviruses such as Rift "Valley fever virus (Anyamba et al., 2002; Linthicum et al., 1991). Longer, interdecadal trends may indicate shifts in baselines (i.e., climate change), and these gradual changes may elongate transmission seasons and extend vector and pathogen distributions. Change has been most clearly detected at northern latitudes (Githeko et al., 2000) and in urban landscapes that present their own microcosms for climate change and variation (Kalnay and Cai, 2003). Certainly the consistently


high incidence of WNV in the U.S. central plains and central Canada (data not shown) was unexpected and generally has tracked above-normal summer temperatures (see Figure 3-4), even though these relatively low mean temperatures were considered suboptimal for virus amplification within the primary mosquito vector Culex tarsalis (Reisen et al., 2006b).

Impact of Climate Variation on Vectors and Pathogens

Vectors and the pathogens they transmit are especially subject to climate variation because climate markedly affects arthropod population size and dynamics, and because pathogen replication rates are influenced directly by ambient temperatures during infection of the poikilothermic10 arthropod vector. This is especially true for the mosquito-borne encephalitides at temperate latitudes where temperature delineates amplification and transmission season duration. Climate variation also indirectly determines the size and age structure of avian maintenance and amplifying host populations by impacting primary productivity and therefore the abundance and distribution of food sources.

In contrast, the impact of climate on mosquito populations is rapid and direct. Many vector mosquitoes utilize surface water accumulations for larval development that depend directly or indirectly on precipitation. For rural species such as Culex tarsalis, the timing and size of the adult population peaks depend directly on winter and spring rainfall and snowmelt runoff, and have been related to El Niño conditions and the associated wet winter and spring seasons (Reisen et al., 2008a). Conversely, urban species, such as those in the Culex pipiens complex, utilize peridomestic and underground drainage systems and may be favored by La Niña conditions of high temperature and low rainfall. In urban centers such as Los Angeles, high rainfall and associated runoff scour the underground larval habitats and actually reduce vector abundance (Su et al., 2003).

Warm temperatures increase the rate of mosquito population growth, reduce adult survival, and increase the frequency of blood feeding as well as the chances of virus acquisition and transmission (Reeves et al., 1994; Reisen, 1995). Temperature also is positively associated with encephalitis virus replication within the mosquito vector. The time from virus infection to transmission (the extrinsic incubation period [EIP]) is directly related to ambient conditions and can be described by degree-day models (Reisen et al., 1993, 2006b). These regimens frequently delineate episodic waves of transmission during outbreaks (Nielsen et al., 2008) as well as transmission seasons and the geographic distribution of virus amplification.

10Any organism whose body temperature varies with the temperature of its surroundings.

FIGURE 3-4 Incidence of human West Nile virus cases per million population and temperature anomalies for the United States, 2003-2007. SOURCES: CDC (2007); NOAA (2008).


Components of a Vector-Borne Pathogen Surveillance Program

Comprehensive vector-borne disease surveillance programs include measures of early-season meteorological conditions that may be used to forecast risk for the coming virus transmission season. As temperatures increase and vectors become active, these programs begin monitoring vector abundance and virus activity within the primary enzootic cycle as a measure of the risk of pathogen transmission to humans and to validate earlier forecasts. These data can be combined in models that forecast or measure risk. For WNV in California, climate measures focus on temperature because the effects of precipitation vary among regions (Reisen et al., 2008a) and vector species (e.g., urban Cx. pipiens mosquitoes do well under hot, dry conditions, whereas rural Cx. tarsalis do well under wet conditions in many areas). Culex abundance is monitored by New Jersey or American light traps (Mulhern, 1985), the Centers for Disease Control and Prevention's (CDC's) dry ice-baited traps (Newhouse et al., 1966), and gravid female traps (Reiter, 1987) in urban situations. Mosquitoes and dead birds are tested rapidly for infection by detecting viral RNA using robotic extraction and real-time reverse transcriptase-polymerase chain reaction (RT-PCR) testing procedures (Shi et al., 2001). Avian infection indicating active transmission is measured by sequentially bleeding sentinel birds, such as chickens, for seroconversion11 and by recording and testing dead birds reported by the public. Equine and human cases are diagnosed by healthcare providers and confirmed serologically at local laboratories. The temporal cascade of events and surveillance data along a typical WNV amplification curve is shown in Figure 3-5. The horizontal line shows the hypothetical level of amplification required before tangential transmission to humans is frequent.

The key to effective surveillance is the rapid dissemination of results and analysis to individuals responsible for intervention decisions. To expedite data exchange in California, a web-based management system called the Surveillance Gateway© has been implemented in which data are entered, stored, analyzed, and displayed (see Figure 3-6). When combined with rapid laboratory diagnostics, mosquitoes collected on Monday or Tuesday and immediately shipped to the laboratory are routinely tested, recorded, reported, and mapped online as early as Thursday or Friday of the same week. These data are then compared to historical records, and a risk score is assigned for each parameter (see Table 3-5). The scores for individual surveillance parameters are then averaged to obtain an overall risk score ranging from 1 to 5, where 1.0-2.5 denotes a "normal season," 2.6-4.0 represents increasing risk requiring "emergency planning," and 4.1-5.0 indicates "epidemic conditions."

Risk levels for WNV can be forecast and then measured based on average daily temperatures (see Figure 3-7). These risk levels, based on temperature,

11Conversion of a sentinel host from antibody negative to positive after infection.

Was this article helpful?

0 0

Post a comment