are not recognized until they are well under way. Thus, predicting where and when epidemics are likely to occur is a key goal of ArboNET.

For clues to WNV behavior in the future, one might look at long-term SLE incidence patterns. Unfortunately, as shown in Figure 2-8, such a pattern is not discernable from 70 years of SLE incidence data. Because WNV has an ecology similar to that of SLE, it is likely that WNV will behave in a similarly unpredictable pattern. However, WNV produces considerably higher levels of viremia in birds, affording it much greater epidemic potential (Komar et al., 2003). Although both WNV and SLE outbreaks, particularly in the northern United States, have often occurred during heat waves, it is noteworthy that the largest U.S. outbreak of SLE was not associated with a heat wave or with any other readily identifiable weather anomaly.

Ecological surveillance can be somewhat helpful in predicting WNV outbreaks. In North America, chickens, mosquitoes, horses, and birds have demonstrated increased activity prior to the onset of, or early in, human outbreaks of WNV illness (Eidson et al., 2001; Kulasekera et al., 2001). However, at best, ecological surveillance provides only a few weeks’ warning before a human WNV outbreak. In Latin America, extensive serological data from ecological surveillance in birds and horses shows that WNV has spread from the Caribbean as far south as Argentina (Morales et al., 2006; Komar and Clark, 2006),

FIGURE 2-8 Human cases of West Nile virus and St. Louis encephalitis neuroinvasive disease, by year, 1932-2006, United States.