Seasonal host switching by mosquitoes may drive arbovirus epizootics

For many mosquito-borne viruses, the classic transmission scenario involves ornithophilic mosquitoes vectoring virus among avian hosts in a cycle of enzootic amplification. Mammals become infected when virus “spills over” into non-avian populations. Despite its importance in virus transmission, the mechanism underlying spill over is not clearly understood. One hypothesis predicts that when a critical percentage of the bird population becomes infected, virus is introduced into non-avian hosts by mosquitoes with catholic feeding habits. A second hypothesis maintains that vectors shift from feeding on birds to feeding upon mammals in an annual pattern of host switching. We tested these hypotheses by quantifying host use of mosquitoes at a wetland study site with documented arbovirus activity. Over a five-year period hosts of mosquitoes were determined by PCR from blood recovered from the guts of 893 field-collected mosquitoes. In each year a shift in host preference was observed for Cx. erraticus, a suspected bridge vector of eastern equine encephalitis virus. Culex erraticus fed more or less equally on birds and mammals in March and April, then shifted to feeding almost entirely on birds in May and June, (sometimes >90%). In July or August of each year mosquitoes switched from feeding on birds to feeding on mammals (often >90%). The annual pattern of host switching was correlated with arbovirus infections in mammals (horses). Our work indicates that host-switching in vector mosquitoes is critically important in initiation of epizootic transmission of mosquito-borne viruses in nature. We hypothesize that the dramatic shifts in host use by mosquitoes is largely driven by host reproductive biology.