Abstract: Risk of human infection with West Nile virus (WNV) is driven by fine-scale vector-host interactions. An intensive field study was conducted in Davis, California to develop models for Cx. tarsalis abundance and WNV prevalence in relation to local spatial features, including the urban edge and nearby rice fields. Bayesian predictive models revealed differing spatial trends between Cx. tarsalis abundance and WNV prevalence, providing practical targets for control efforts based on vector abundance or prevalence thresholds.
