Impact of climate on western equine encephalitis in Manitoba, Minnesota and North Dakota, 1980-1983.

Information was collected on confirmed outbreaks of western equine encephalitis (WEE) in North America east of the Rockies for 1981 and 1983 (epidemic years) and 1980 and 1982 (non-epidemic years). The initial pattern of outbreaks in Manitoba, Minnesota and North Dakota was determined for each year. Backward (and in some instances forward) wind trajectories were computed for each day 4-15 days (incubation period) before the initial outbreaks of WEE in a given area of province or state. During these years the timing and location of WEE outbreaks in horses and man, seroconversion in chickens, the maximum Culex tarsalis counts at Winnipeg and first isolation of WEE virus from C. tarsalis could be correlated with trajectories of winds from states further south within acceptable intervals. It is suggested that C. tarsalis mosquitoes infected with WEE virus are carried on the wind from Texas on the Gulf of Mexico, where they continue to breed during the northern winter months, to northern Texas and Oklahoma in the spring. In May, June and July C. tarsalis are carried north on southerly winds from these states through Kansas and Nebraska to North Dakota, Minnesota, Wisconsin and Manitoba. Distances of 1250-1350 km are traversed in 18-24 h at heights up to 1.5 km with temperatures greater than or equal to 13 degrees C. Landing takes place where the warm southerly winds meet cold fronts associated with rain. Convergence leads to concentration of C. tarsalis and determines where outbreaks occur. It is possible that return of new generations of C. tarsalis to the south may occur later in the year. The development of an epidemic of WEE in the northern states and provinces would appear to depend on (i) suitable trajectories from the south in June and July with temperatures greater than or equal to 13 degrees C meeting cold fronts with rain, (ii) sufficient C. tarsalis infected with WEE virus at source, carried on the wind and locally, (iii) C. tarsalis biting horses and man, (iv) maintenance of local mosquito populations in August and (v) susceptible hosts (birds) at source and susceptible hosts (horses and man) locally. Possible methods of prediction involving determination of trajectories, identification of C. tarsalis blood meals, measuring seroconversion in calves are discussed in addition to the methods already in use.