One of the most important problems in understanding arthropod-transmitted diseases is to explain how many pathogenic agents maintain transmission at their seemingly low abundance and at the low rate of survival established for their insect vectors. We are investigating this topic by determining if vertebrates infected with arboviruses are fed upon more often by mosquitoes than are uninfected hosts. If this difference occurs, the probability of mosquito infection may be higher than currently is predicted.

Experiments on the effect of St. Louis encephalitis, western equine encephalomyelitis, or eastern equine encephalomyelitis viruses on antimosquito behavior of adult and nestling house sparrows (Passer domesticus) fail to support the increased feeding hypothesis. Adult birds were defensive and infection did not reduce their ability to repel mosquitoes. Nestlings were not defensive, so although they were more susceptible to clinical illness than adults, infection could not modify their behavior and increase mosquito feeding. However, a critical period for the development of nestling antimosquito behavior was observed. We suggest that the association of that critical period with the corresponding increased susceptibility to disease deserves additional study. Overall the factors that appear to have the most potentially important effect on increased mosquito feeding success are the type of virus inoculated, viral dose, and host age. Two other mechanisms that could increase mosquito feeding success on an infected host also need to be examined. They are host attraction and the ease with which probing mosquitoes can locate blood.